Download Report - Institute for Clinical and Economic Review

Document related concepts

Patient safety wikipedia , lookup

Gene therapy wikipedia , lookup

Clinical trial wikipedia , lookup

Adherence (medicine) wikipedia , lookup

Placebo-controlled study wikipedia , lookup

Alzheimer's disease research wikipedia , lookup

Management of multiple sclerosis wikipedia , lookup

Multiple sclerosis research wikipedia , lookup

Transcript
PCSK9 Inhibitors for Treatment of High
Cholesterol: Effectiveness, Value, and ValueBased Price Benchmarks
Final Report
November 24, 2015
Completed by:
Institute for Clinical and Economic Review
©Institute for Clinical and Economic Review, 2015
AUTHORS
ICER Staff & Affiliated Researchers
Jeffrey A. Tice, MD
Associate Professor of Medicine, Division of
General Internal Medicine, Department of
Medicine, University of California San Francisco
Daniel A. Ollendorf, PhD
Chief Review Officer
Institute for Clinical and Economic Review
Courtney Cunningham, MPH
Program Director
Institute for Clinical and Economic Review
Steven D. Pearson, MD, MSc
President
Institute for Clinical and Economic Review
CVD Policy Model Group
Dhruv S. Kazi, MD, MSc, MS
Assistant Professor
Department of Medicine (Cardiology), Department of
Epidemiology and Biostatistics, and Center for
Healthcare Value, University of California San Francisco
Pamela G. Coxson, PhD
Mathematics Specialist
Department of Medicine, University of California San
Francisco
Andrew E. Moran, MD
Assistant Professor
Division of General Internal Medicine, Department of
Medicine, Columbia University
Joanne Penko, MS, MPH
Research Analyst
Center for Vulnerable Populations, Department of
Medicine, University of California San Francisco
David Guzman, MSPH
Biostatistician
Center for Vulnerable Populations, Department of
Medicine, University of California San Francisco
Kirsten Bibbins-Domingo, MD, PhD, MAS
Professor
Department of Medicine, Department of Epidemiology
and Biostatistics, and Center for Vulnerable
Populations, University of California San Francisco
DATE OF
PUBLICATION:
November 24, 2015
We would also like to thank Erin Lawler and Matt Seidner of ICER for their contributions to this
report.
©Institute for Clinical and Economic Review, 2015
Page i
Table of Contents
Executive Summary ............................................................................................................................ ES1
Background .................................................................................................................................... ES1
Topic in Context ............................................................................................................................. ES1
Comparative Clinical Effectiveness ................................................................................................ ES4
Comparative Value: Incremental Costs per Outcomes Achieved .................................................. ES9
Comparative Value: Health System Value ................................................................................... ES15
Draft Value-based Benchmark Prices .......................................................................................... ES17
1. Background ........................................................................................................................................ 1
1.1 Introduction ................................................................................................................................. 1
2. The Topic in Context .......................................................................................................................... 4
3. Summary of Coverage Policies ........................................................................................................... 9
3.1 Summary of Coverage for PCSK9 inhibitors: Praluent and Repatha ............................................ 9
3.2 Summary of Coverage for Existing Lipid-Lowering Therapies ................................................... 12
4. Comparative Clinical Effectiveness .................................................................................................. 14
4.1 Methods ..................................................................................................................................... 14
4.2 Results ........................................................................................................................................ 17
4.3 Summary and Comment ............................................................................................................ 25
5. Other Benefits or Disadvantages ..................................................................................................... 27
6. Comparative Value ........................................................................................................................... 28
6.1 Overview .................................................................................................................................... 28
6.2 Incremental Costs per Outcomes Achieved ............................................................................... 28
6.3 Health System Value .................................................................................................................. 47
6.4 Draft Value-based Benchmark Prices ........................................................................................ 51
6.5 Summary and Comment ............................................................................................................ 52
7. Summary of the Votes and Considerations for Policy ..................................................................... 56
7.1 About the New England CEPAC Process .................................................................................... 56
7.2 Comparative Clinical Effectiveness Voting Results .................................................................... 60
7.3 Care Value Voting Results .......................................................................................................... 61
7.4 Provisional Health System Value Voting Results ....................................................................... 62
©Institute for Clinical and Economic Review, 2015
Page ii
7.5 Roundtable Discussion and Key Policy Recommendations ....................................................... 64
References ........................................................................................................................................... 70
A1. Search Strategies ........................................................................................................................... 82
A2. Clinical Guidelines .......................................................................................................................... 84
A3. Detailed Coverage Policies ............................................................................................................ 90
A4. Previous Systematic Reviews and Technology Assessments......................................................... 92
A5. Ongoing Studies ............................................................................................................................. 93
A6. Comparative Clinical Effectiveness Appendix ................................................................................ 94
A7. Comparative Value Appendix ........................................................................................................ 99
A8: Meeting Agenda and Participants ............................................................................................... 114
©Institute for Clinical and Economic Review, 2015
Page iii
List of Abbreviations Used in this Report
AACE
American Association of Clinical Endocrinologists
ACC
American College of Cardiology
AHA
American Heart Association
AHRQ
Agency for Healthcare Research and Quality
ASCVD
Atherosclerotic cardiovascular disease
AE
Adverse event
CHD
Coronary heart disease
CHF
Congestive heart failure
CK
Creatinine kinase
CVD
Cardiovascular disease
EAS
European Atherosclerosis Society
ESC
European Society of Cardiology
FH
Familial hypercholesterolemia
HeFH
Heterozygous familial hypercholesterolemia
HC
Non-specific hypercholesterolemia
HoFH
Homozygous familial hypercholesterolemia
ICER
Incremental cost effectiveness ratio
LDL-C
Low-density lipoprotein cholesterol
MACE
Major adverse cardiovascular event
MI
Myocardial infarction
NICE
National Institute for Health and Care Excellence
NNT
Number needed to treat
PCSK9
Proprotein convertase subtilisin/kexin type 9
QALY
Quality-adjusted life year
SBP
Systolic blood pressure
©Institute for Clinical and Economic Review, 2015
Page iv
About ICER
The Institute for Clinical and Economic Review (ICER) is an independent non-profit research
organization that evaluates medical evidence and convenes public deliberative bodies to help
stakeholders interpret and apply evidence to improve patient outcomes and control costs. ICER
receives funding from government grants, non-profit foundations, health plans, provider groups,
and health industry manufacturers. Through all its work, ICER seeks to help create a future in which
collaborative efforts to move evidence into action provide the foundation for a more effective,
efficient, and just health care system. More information about ICER is available at http://www.icerreview.org
About New England CEPAC
The New England Comparative Effectiveness Public Advisory Council (CEPAC) is an independent,
regional body of practicing physicians, methodological experts, and leaders in patient advocacy and
engagement that provides objective, independent guidance on the application of medical evidence
to clinical practice and payer policy decisions across New England.
Council members are selected for three-year terms, and represent a diversity of expertise and
perspective; they are purposely not selected for expertise in the clinical topic under discussion in
order to maintain the objectivity of the Council and to ground the conversation in the interpretation
of the published evidence rather than anecdotal experience or expert opinion. Acknowledging that
any judgment of evidence is strengthened by real life clinical and patient perspectives, New England
CEPAC recruits subject matter experts for each meeting who provide input to Council members
before the meeting to help clarify New England CEPAC’s understanding of the different
interventions being analyzed in the evidence review. The same clinical experts serve as a resource
to the Council during their public deliberation, and help form recommendations with the New
England CEPAC on ways the evidence can be applied to policy and practice.
Led by the Institute for Clinical and Economic Review, the New England CEPAC is supported by a
broad coalition of state Medicaid leaders, integrated provider groups, public and private payers,
patient representatives, and philanthropy. For more information about the New England CEPAC,
please visit cepac.icer-review.org.
©Institute for Clinical and Economic Review, 2015
Page v
Executive Summary
Background
The goal of this report is to address the key issues that patients, providers, and payers face when
making decisions about PCSK9 inhibitor therapy and to support the dialogue needed for successful
action to improve the quality and value of health care for all patients.
The scope for this assessment utilizes the PICOTS (Population, Intervention, Comparators,
Outcomes, Timing, and Settings) framework. The evidence review is based on 25 clinical trials and
two published systematic reviews and meta-analyses.1,2 The results were cross-checked with the
manufacturers’ FDA submission documents and the FDA’s briefing documents.
Topic in Context
In summarizing the contextual considerations for appraisal of a health care intervention, we seek to
highlight the four following specific issues:

Is there a particularly high burden/severity of illness?

Do other acceptable treatments exist?

Are other, equally or more effective treatments nearing introduction into practice?

Would other societal values accord substantially more or less priority to providing access to
this treatment for this patient population?
Elevated Cholesterol, Statin Therapy, and Cardiovascular Outcomes
Approximately one-third of American adults have cardiovascular disease (CVD), making it the most
common cause of death in the United States.3 The American Heart Association (AHA) defines
cardiovascular disease as anyone with a history of myocardial infarction (MI), stroke, angina,
congestive heart failure (CHF), peripheral artery disease, or hypertension. Major cardiovascular
events in clinical trials usually include death due to CVD, non-fatal MI, non-fatal stroke, unstable
angina requiring hospitalization, and revascularization (stenting, bypass surgery). Biological and
epidemiological evidence has linked high levels of circulating low-density lipoprotein cholesterol
(LDL-C) with an increased risk of MI, stroke, and death from CVD.
Multiple randomized clinical trials have demonstrated that lowering LDL-C with statin therapy
reduces the risk of MI, stroke, and death from CVD. Many investigators believe that the greater the
reduction in LDL -C the greater the reduction in cardiovascular events, but the topic remains
controversial. However, several drugs that lower LDL-C – including hormone therapy, niacin, and
©Institute for Clinical and Economic Review, 2015
Page ES1
torcetrapib – have not decreased cardiovascular disease events when evaluated in randomized
trials despite lowering LDL-C. On the other hand, the recently published IMPROVE-IT trial
demonstrated that lowering LDL-C with ezetimibe significantly reduced cardiovascular event rates
by 6% (95% CI 1 to 11%) after a median follow-up of approximately 5 years.
Guidelines for Cholesterol Lowering Therapy
In 2013 the ACC/AHA released an updated guideline for the treatment of cholesterol in order to
reduce cardiovascular risk in adults.4 The guideline includes a “strong” recommendation for high
intensity statin therapy to treat individuals with cardiovascular disease who are ≤ 75 years of age;
moderate intensity statin use in individuals with diabetes mellitus and LDL-C levels between 70 and
189 mg/dL who are ages 40-75 years of age; and high intensity statin use in individuals aged 40-75
with a 10-year risk for cardiovascular disease ≥ 7.5% and LDL-C levels between 70 and 189 mg/dL.
The guideline also makes a “moderate” recommendation for high intensity statin therapy to treat
all individuals with LDL-C levels ≥ 190 mg/dL who are ≥ 21 years of age.
The major change in this 2013 ACC/AHA guideline compared to the earlier National Cholesterol
Education Program Adult Treatment Panel III (NCEP ATP III) guideline was its move away from
recommending specific LDL-C levels as treatment targets.5 In the prior guidelines, statin therapy
was recommended to reach a target LDL-C level of < 100 mg/dL for individuals with cardiovascular
disease and those with a 10-year risk ≥ 20%. For individuals with multiple risk factors and a 10-year
risk < 20%, the target LDL-C level was < 130 mg/dL. The 2011 European guidelines also recommend
statin therapy to reach a target LDL-C level of < 70 mg/dL for individuals with cardiovascular disease
or diabetes and < 100 mg/dL for primary prevention in high risk individuals.6 The emergence of the
PCSK9 inhibitor drugs has stimulated further debate about whether clinicians should seek to
achieve specific LDL-C targets when treating different kinds of patients.7
Unmet Clinical Need
Patient populations with elevated cholesterol in which there is an unmet clinical need include
patients with a genetic condition causing highly elevated LDL-C, patients on statins and/or other
cholesterol lowering drugs who are felt to have had an inadequate reduction in LDL-C, and patients
who are not able to tolerate statins.8
Familial hypercholesterolemia (FH)
Familial hypercholesterolemia (FH) is an autosomal dominant inherited condition that causes
elevated LDL-C in both the heterozygous (HeFH) and homozygous (HoFH) states.9-12 Individuals with
HoFH have LDL-C levels > 500 mg/dL and often experience cardiovascular events by age 20. It is an
extremely rare condition (~1 case per 1 million people) with only 300-400 individuals in the US
affected by HoFH. HeFH is more common, with estimates of affected individuals in the US varying
between 500,000 to over a million, although there is considerable uncertainty in this estimate.
©Institute for Clinical and Economic Review, 2015
Page ES2
Individuals with HeFH often have LDL-C levels that are two to three times normal (i.e., 250-350
mg/dL).13,14
Statin intolerance
Statin intolerance is primarily due to muscle symptoms. These range from asymptomatic mild
elevations in creatinine kinase (CK), a muscle enzyme, to muscle aches (myalgias) with or without
mild elevations in CK (<4 times the upper limit of normal), to frank myositis (CK ≥ 4 time the upper
limit of normal).15,16 Precise measurement of statin intolerance is difficult because muscle
symptoms arising from other causes are common, particularly in older individuals. Two studies
specifically examined statin intolerance in clinical practice. The Prediction of Muscular Risk in
Observational (PRIMO) trial reported a 10% incidence of mild to moderate muscle symptoms for
patients on high intensity statin therapy.17 Similarly, the Effect of Statins on Skeletal Muscle
Function and Performance (STOMP) study reported a 9.4% incidence of muscle symptoms in statinnaïve patients treated with atorvastatin 80 mg daily compared to a 4.6% incidence in patients
randomized to placebo.18
Other Drug Therapy Options: Ezetimibe (Zetia®)
For patients unable to take statins, or for patients taking statins but not meeting their LDL-C goals,
additional drugs are available, with ezetimibe (Zetia®) being the most relevant for this review.
Ezetimibe inhibits the absorption of cholesterol in the intestines. Meta-analyses of randomized
trials suggest that ezetimibe 10 mg lowers LDL-C by 23.6% (95% CI 21.7 to 25.6%) when added to
statin therapy and by 18.6% (95% CI 17.5 to 19.7%) as monotherapy.19,20 However, treatment with
ezetimibe has been controversial because of negative findings in two trials.21,22 These were small
trials that were not designed to evaluate cardiovascular outcomes. The more recent IMPROVE-IT
trial randomized 18,144 patients hospitalized for an acute coronary syndrome in the prior 10 days
to the combination of simvastatin and ezetimibe or simvastatin and placebo and followed them for
a median of approximately 5 years. The estimated cumulative event rate at 7 years was 32.7% in
the ezetimibe group and 34.7% in the placebo group (p=0.016). The publication of the IMPROVE-IT
trial in 2015 has renewed enthusiasm among many cardiologists for the use of ezetimibe to lower
LDL-C beyond the reduction achieved with statin therapy.
Proprotein convertase subtilisin/kexin type 9 (PCSK9) and cardiovascular disease
Higher levels of PCSK9 reduce the number of LDL-C receptors. If there are fewer LDL-C receptors,
then LDL-C levels rise in the blood. Conversely, lower levels of PCSK9 in the blood leads higher LDL-C
receptor density and lower levels of LDL-C in the blood. This biology suggests that drugs targeting
PCSK9 have the potential to reduce LDL-C and cardiovascular disease.
In July and August 2015, after favorable votes from its Advisory Committee ranging from 11-4 to 150 for different indications, the FDA approved two new human monoclonal antibodies that target
©Institute for Clinical and Economic Review, 2015
Page ES3
PCSK9 in the blood and markedly reduce LDL -C levels. Alirocumab (Praluent®, Sanofi/Regeneron) is
a human monoclonal antibody that inhibits PCSK9. It is administered as a subcutaneous injection
once every two weeks at doses of either 75 mg or 150 mg. Evolocumab (Repatha™, Amgen) is also a
human monoclonal antibody and is administered as a subcutaneous injection 140 mg once every
two weeks or 420 mg once every four weeks. The annual wholesale acquisition cost for treatment
with alirocumab is $14,600; the annual cost for evolocumab is $14,100.
Comparative Clinical Effectiveness
Methods
The goal of this report is to evaluate the comparative clinical effectiveness and comparative value of
PCSK9 inhibitors as a class for patients with elevated LDL-C. We have attempted to identify all
randomized controlled trials that evaluated the safety and efficacy of the two FDA approved PCSK9
inhibitors alirocumab and evolocumab. The published meta-analyses found that the LDL-C lowering
effect of the two PCSK9 inhibitors were similar, and there are no head to head trials that compare
alirocumab to evolocumab; thus, we have elected to examine the impact of PCSK9 inhibitors as a
class.
Results
Our literature search identified 41 references describing eight phase 2 trials, 16 phase 3 trials, and
one long-term follow-up study.8,23-62 A high-quality meta-analysis by Navarese and colleagues was
also identified and provided the basis for many of the findings in this review.1 Most of the clinical
trials were of relatively short duration. Seventeen trials had follow-up of <1 year, two trials had one
year of follow-up, and five trials had follow-up longer than one year. Fourteen trials involved
comparisons of PCSK9 inhibitors to placebo, seven compared PSCK9 inhibitors to ezetimibe, and
three involved both comparisons. Approximately equal percentages of trial participants were male
and female, 30% had a history of CVD, and 15% had diabetes. Key trials are summarized in detail in
the full report.
Clinical Benefits
LDL-C reduction and other lipid parameters
Table ES1 on the following page shows the results of the Navarese meta-analysis and demonstrates
that the clinical impact on LDL-C is very similar between the two drugs. Evolocumab has slightly
greater LDL-C reductions than alirocumab, but the differences are very small compared to the
percentage reduction achieved by either of the PCSK9 inhibitors. Furthermore, differences in the
underlying populations studied may explain these relatively small differences. The evidence
therefore strongly suggests that the two drugs have very similar effects, and the lack of head to
©Institute for Clinical and Economic Review, 2015
Page ES4
head randomized trials makes it impossible to determine whether one of the PCSK9 inhibitors
lowers cholesterol more than the other.
Table ES1. Meta-analysis of the percentage reduction in LDL-C by PCSK9 inhibitors in 10,159
participants in phase 2 and 3 randomized trials by stratified by dose and type of PCSK9 inhibitor.
Dose and type of PCSK9 inhibitor
Comparison group
All, % (95% CI)
Alirocumab 75
mg Q2W
Alirocumab
150 mg Q2W
Evolocumab
140 mg Q2W
Evolocumab
420 mg Q4W
Placebo
58.8 (56.5 to 61.0)
52.6
56.2
63.5
57.3
Ezetimibe
36.2 (33.1 to 39.3)
31.7
*
39.3
37.5
Given these findings, the Navarese meta-analysis merged data available on both drugs to evaluate
their impact on other lipid parameters.
Table ES2. Meta-analysis of the percentage reduction in LDL-C by PCSK9 inhibitors as a class in
10,159 participants in phase 2 and 3 randomized trials stratified by background statin therapy.
Comparison group
Placebo
Ezetimibe
All, % (95% CI)
58.8 (56.5 to 61.0)
36.2 (33.1 to 39.3)
Background Statin Therapy
No statin, %
High intensity
statin, %
53.6
57.9
36.2
34.4
Other statin, %
65.2
37.5
The percentage reduction in LDL is greater when PCSK9 inhibitors are compared to placebo (58.8%)
than that observed compared to ezetimibe (36.2%). The percentage reduction in LDL varies much
less by background statin therapy. Detailed information on LDL-lowering by subgroup is presented
in the full report. Findings from studies conducted in patients with HeFH and statin intolerance
were similar to the overall results.
Patient-centered clinical outcomes
There are 5-year large outcome studies ongoing for both alirocumab and evolocumab that should
present initial results in 2017. Individual studies completed to date were not powered to evaluate
outcomes such as mortality or CVD adverse events. However, the meta-analysis by Navarese
combined data from existing studies to examine these outcomes. The most important clinical
outcomes for lipid lowering therapy include death from CVD, MI, stroke, and unstable angina
requiring hospitalization. Navarese and colleagues did not report the stroke outcomes, so we
performed our own meta-analysis of stroke outcomes using the same analytic methods (see Table
ES3 on the next page).
©Institute for Clinical and Economic Review, 2015
Page ES5
Table ES3: Meta-analysis results for patient-oriented outcomes
Outcome
All-cause
mortality
CVD Mortality
MI
Stroke
Unstable angina
OR (95% CI)
P
I2
N
Events PCSK9 group (%)
0.45 (0.23-0.86)
0.015
0%
10,159
19 (0.3%)
Events control
group (%)
21 (0.5%)
0.50 (0.23-1.10)
0.49 (0.26-0.93)
1.97 (0.69-5.65)
0.61 (0.06-6.14)
0.084
0.030
0.206
0.676
0%
0%
0%
0%
10,159
5,195
4,683
3,894
12 (0.2%)
19 (0.6%)
14 (0.5%)
1 (0.05%)
13 (0.3%)
19 (1.0%)
3 (0.2%)
1 (0.08%)
As shown in the table above, the findings of the meta-analysis suggest that the PCSK9 inhibitors
reduce the odds of all-cause and cardiovascular mortality by about 50%, but the total number of
events is low and the confidence intervals are wide. The odds ratio for stroke in the meta-analysis
was twice as high in the PCSK9 group, but the confidence interval is very wide and not statistically
significant. There were no significant differences in these results when stratified by comparison
group (placebo, ezetimibe), by PCSK9 inhibitor (alirocumab, evolocumab) or when adjusted for
length of follow-up. In sensitivity analyses, excluding the data from studies not yet published in the
peer-reviewed literature, the conclusions are the same.
Harms
Nearly all studies have less than 6 months of follow-up data, but results from individual studies and
from the Navarese meta-analysis have found that PCSK9 drugs are very well-tolerated; there have
been no findings suggestive of significant increases in adverse event rates. There are more injection
site reactions, which may lead to slightly higher rates of drug discontinuation compared to the
control group. There is a slight excess of neurocognitive events with PCSK9 inhibitors, but the
results are not statistically significant. There is also a trend towards more myalgias in the PCSK9
treated participants, but this is balanced by a statistically significant reduction in the number of
participants with elevations in the muscle enzyme creatine kinase (CK). Detailed adverse event-rate
data are provided in the full report.
Summary and Comment
Our analyses demonstrate that the existing evidence provides moderate certainty that PCSK9
treatment provides an incremental or substantial net health benefit for all of the patient
subpopulations included in the scope of this review. There is no question that the drugs improve
intermediate risk factors for cardiovascular disease. They substantially reduce LDL-C, total
cholesterol, and lipoprotein(a), and also modestly elevate HDL-C. A high-quality meta-analysis
found a 50% reduction in all-cause mortality that was statistically significant and reductions of
similar magnitude (albeit not statistically significant) in death from cardiovascular disease and in
MIs.
©Institute for Clinical and Economic Review, 2015
Page ES6
The drugs also appear to be very well-tolerated. The randomized trials do not demonstrate an
increase in adverse events, serious adverse events, or drug discontinuations due to adverse events.
Neurocognitive event rates are low and do not appear to be increased in patients randomized to
PCSK9 inhibitors compared to the control patients.
However, there are several limitations in the evidence base that give reason for caution. There are
theoretical concerns that long term exposure to very low levels of cholesterol may have unexpected
adverse effects that have not been observed in the evidence base to date because the majority of
the studies lasted less than six months. In addition, as noted earlier, medications such as
torcetrapib that lower LDL-C, raise HDL, and have strong biological plausibility, have demonstrated
in long term studies increased cardiovascular event rates and total mortality. The large randomized
trials with long-term follow-up that are designed to evaluate the effect of the PCSK9 inhibitors on
hard clinical endpoints have completed recruitment, but their results will not be available until
2017.
The promising evidence on patient-centered outcomes from the published meta-analysis is also
limited in several ways. First, the 95% confidence intervals for the odds ratios estimating clinical
benefit either include 1.0 or approach 1.0. Second, the evidence in this meta-analysis combines
data from trials of two different PCSK9 inhibitors, each with two different dosing schedules, with
too few events in the evidence base to attempt subgroup analyses. Another limitation of the metaanalysis is that the populations studied were quite different: young adults with homozygous FH and
very high LDL-C; older adults with LDL-C < 100, but not at goal; and older adults who have already
had a heart attack or stroke. A final reason for caution about the findings of the meta-analysis is
that the PCSK9 inhibitors were compared to two different control arms: placebo and ezetimibe. The
percentage LDL-C reduction consistently favored PCSK9 inhibitors, but the magnitude varied slightly
by population and significantly by control group. It is likely that the clinical benefits will vary by
dose, drug, background drug therapy, and population studied.
The evidence base provides high certainty, however, that PCSK9 inhibitors lead to superior
reductions in LDL-C levels compared to both placebo and ezetimibe. The percent reduction in LDL-C
with PCKS9 treatment is approximately 55-60% and appears not to differ substantially across
different patient subpopulations. The potential net health benefit from this level of LDL-C reduction
will be greater among patient subpopulations at higher risks of CVD. Among the subgroups, the
population with HoFH is at highest risk for CVD events. Untreated, they have CVD events in the
second decade of life. Differences in CVD risk are less marked between patients with HeFH and
those with a prior history of CVD who have elevated LDL-C levels despite other treatment and/or
who cannot take statins.
In summary, the ICER review team believes that the existing evidence suggests, with moderate
certainty, that the net health benefit of the PCSK9 inhibitors is either incremental or substantial for
the patients in the subpopulations within the scope of this review. Despite the uncertainty in the
©Institute for Clinical and Economic Review, 2015
Page ES7
actual level of net health benefit, we believe there is less than a 10% chance that ongoing trials will
demonstrate a net harm from PCSK9 inhibitor treatment, and therefore our evidence rating within
the ICER Integrated Evidence Rating framework is “Promising but Inconclusive.”
Other Benefits or Disadvantages
Our reviews seek to provide information on other benefits or disadvantages offered by the
intervention to the individual patient, caregivers, the delivery system, other patients, or the public
that would not have been considered as part of the evidence on comparative clinical effectiveness.
Examples include, but are not limited to:

Methods of administration that improve or diminish patient acceptability and adherence

A public health benefit, e.g. reducing new infections

Treatment outcomes that reduce disparities across various patient groups

More rapid return to work or other positive effects on productivity (if not considered a
benefit as part of comparative clinical effectiveness)

New mechanisms of action for treatments of clinical conditions (e.g., mental illness) for
which the response to currently available treatments varies significantly among patients for
unknown reasons (substantial heterogeneity of treatment effect)
Currently available PCSK9 inhibitors must be injected. This is a potential disadvantage compared to
most pharmaceuticals because some patients are unable to self-inject or experience anxiety
associated with self-injection. On the other hand, patients rapidly learn to inject themselves with
low molecular weight heparin and with insulin when needed, so the barrier may not be too high,
particularly for patients motivated by FH or a history of CVD events. Furthermore, the need to inject
the medication only once or twice a month may enhance adherence and be an advantage
compared to medications that need to be taken on a daily basis.
There do not appear to be other benefits or disadvantages of note to PCSK9 inhibitor therapy.
©Institute for Clinical and Economic Review, 2015
Page ES8
Comparative Value: Incremental Costs per Outcomes Achieved
Overview
To assess the incremental costs per outcomes achieved of PCSK9 inhibitors, we conducted a costeffectiveness analysis using the CVD Policy Model, a previously validated model of cardiovascular
disease in the contemporary adult population of the United States. The CVD Policy Model is a
computer-simulation, discrete-state Markov model of coronary heart disease and stroke incidence,
prevalence, mortality, and costs in the U.S. population over age 35 years.63-65 The model was
created at Harvard University in 1984 and has been used for more than 30 years to provide
evidence on the value of cardiovascular disease prevention approaches in U.S. adults. The CVD
Policy Model team has published reports from a number of high-impact studies of public health and
clinical interventions.66-75 The last model software and input data update was completed in 2015.
For the purpose of this analysis, we estimated the degree of LDL-C reduction with PCSK9 inhibitors
when used alone or in combination with statins. We assumed that the drugs were equally
efficacious in all patient populations, i.e., the proportion of reduction in LDL-C from baseline was
constant across all subgroups studied. We also estimated the LDL-lowering effect of ezetimibe,
another second-line LDL-lowering drug, alone or in combination with statin therapy.
We assumed that the effect of these drugs on cardiovascular outcomes (non-fatal MI, stroke and
cardiovascular death) is proportionate to the degree of reduction in LDL-C: for one unit decline in
LDL-C, we assumed that statins, ezetimibe, and PCSK9 inhibitors reduce the risk of non-fatal MI,
non-fatal stroke, and cardiovascular death by an identical amount. Since the effect of PCSK9
inhibitors on stroke is not known, we performed a sensitivity analysis that assumed no change in
the risk of stroke among patients treated with PCSK9 inhibitors.
Cost-Effectiveness Model: Methods
Model Structure
We modeled the entire population of US adults aged 35 to 74 years in the year 2015. We assumed
the health system perspective,76 considering all direct and induced medical costs and relevant
clinical outcomes. In the base case, we evaluated the cost-effectiveness of PCSK9 inhibitors in three
target populations. Populations were chosen to approximate those described in the FDA-labeled
indications for alirocumab (i.e., FH and patients with atherosclerotic cardiovascular disease
(ASCVD)).77 This is in line with the idea that because statins are both inexpensive and effective,
PCSK9 inhibitors will probably be used first among patients at highest risk for adverse
cardiovascular events. Because the available data sources for the model have no variables for
clinically-confirmed FH, we defined this condition based on the presence of a very high LDL-C (>250
mg/dL in the absence of statin use, ≥200mg/dL with statin use). Patients with a history of CVD were
stratified into those intolerant to statins (10% of the overall population) and those on statin therapy
©Institute for Clinical and Economic Review, 2015
Page ES9
but not at LDL-C goal (<70 mg/dL). We applied a lifetime analytic horizon, defined as until patients
reach 95 years of age (because of the absence of high-quality epidemiologic data in older
populations), discounting future costs and benefits by 3% a year.
Treatment Strategies
We modeled three treatment strategies in patients able to tolerate statins:

background treatment with a statin (as treated in the population, control),

incremental treatment with ezetimibe among patients already on a statin, or

incremental treatment with a PCSK9 inhibitor among patients already on a statin.
In the base case, 10% of the population was deemed statin-intolerant. Where relevant, the
treatment strategies available to these patients were:

no treatment with lipid lowering therapies (control),

treatment with ezetimibe, or

treatment with a PCSK9 inhibitor.
In all cases, we assumed that these drugs affect cardiovascular outcomes (non-fatal MI, non-fatal
stroke, and cardiovascular death) in proportion to their effect on LDL-C: for one unit decline in LDLC, we assumed that statins, ezetimibe, and PCSK9 inhibitors reduce the risk of non-fatal MI and
cardiovascular death by an identical amount.
Costs
Age- and sex-specific health care costs were estimated using national data.71 We assumed the
annual cost of ezetimibe to be $2,828, based on the wholesale acquisition cost.78 We assumed the
annual cost of PCSK9 inhibitors to be equal to the average of the recently announced annual
wholesale price of alirocumab and evolocumab ($14,600 and 14,100 per patient per year,
respectively).25 Drug costs were subjected to a variety of sensitivity and threshold analyses.
©Institute for Clinical and Economic Review, 2015
Page ES10
Results
Familial Hypercholesterolemia
Table ES4 below demonstrates that, compared with the control arm, incremental treatment with
ezetimibe would avert 115,900 Major Adverse Cardiac Events (MACE) over the lifetime horizon and
produce 250,600 additional QALYs with an ICER of $135,000/QALY vs. current treatment. Adding
PCSK9 inhibitors to current treatment averted 324,200 MACE and produced 665,200 additional
QALYs, producing an ICER of $290,000/QALY. This higher ICER for PCSK9 inhibitors was driven
largely by differences in drug costs ($14,350 per year for PCSK9 vs. $2,828 per year for ezetimibe).
We did not model HoFH separately, because the expected number of patients is small (n=300-400
in the US).
Table ES4. Base Case and Clinical Outcomes among Patients with FH.
Personyears of
treatment
(millions)
Total
MACE
averted
NNT5†
Statin§
QALYs
gained⌃
Incremental
Drug Costs⌃
(million $)
Incremental
Costs, Other
CV Care⌃
(million $)
ICER
($/QALY)
comparator
Statin +
Ezetimibe||,¶
22.3
115,900
77
250,600
$40,359
-$6,632
$135,000
Statin + PCSK9
inhibitor**,¶
23.7
324,200
28
665,200
$210,516
-$17,304
$290,000
Abbreviations: CV, cardiovascular; FH, familial hypercholesterolemia; ICER, incremental cost-effectiveness ratio; MACE, major
adverse cardiovascular event (nonfatal MI, nonfatal stroke, and cardiovascular death); NNT5, number-needed-to-treat; QALY,
quality-adjusted life year.
* In the base case, all patients who met the operational definition of FH and were either already receiving statin therapy or
deemed statin-intolerant (10% of the population) received incremental therapy with ezetimibe or a PCSK9 inhibitor (n =
605,000 in 2015). The analytic horizon was lifetime (defined as when patients reach the age of 95 years). To reflect the
precision of the model, person-years of treatment are rounded to the nearest 100,000s; MACE and QALYs are rounded to the
100s; costs are rounded to the millions; and ICERs to the 1000s.
† Number of patients that would need to be treated for 5 years to avert one MACE event.
⌃ All costs are reported in 2015 U.S. dollars. Future costs and QALYs are discounted 3% a year.
§ Patients deemed to be statin-intolerant (base-case prevalence = 10% of the FH population) received no lipid-lowering
therapy.
|| Patients deemed to be statin-intolerant (base-case prevalence = 10% of the FH population) received only ezetimibe.
¶ Both statin+ezetimibe and statin+PSCK9 inhibitor arms are compared with the statin-only arm.
** Patients deemed to be statin-intolerant (base-case prevalence = 10% of the FH population) received only a PCSK9 inhibitor.
©Institute for Clinical and Economic Review, 2015
Page ES11
Secondary Prevention Among Patients with a Prior History of CVD and Intolerant of Statins
As shown in the table below, compared with the control arm (no lipid-lowering therapy), treatment
with PCSK9 inhibitors averted 1,254,400 MACE over the lifetime horizon and produced 2,366,000
additional QALYs at an ICER of $274,000/QALY. As in the FH population, ezetimibe’s clinical effects
were less pronounced but its incremental drug costs were approximately 20% of those for PCSK9
inhibitors, resulting in an ICER of $145,000/QALY vs. no lipid-lowering therapy.
Table ES5. Base-Case Clinical and Economic Outcomes Among Statin-Intolerant Patients with a
Prior History of CVD.*
Personyears of
treatment
(millions)
Total
MACE
averted
NNT5†
QALYs
gained⌃
Incremental
Drug Costs⌃
(million $)
Incremental
Costs, Other
CV Care⌃
(million $)
ICER
($/QALY)
Control
comparator
(no additional lipidlowering therapy)
Ezetimibe§
85.0
446,100
56
847,000
$138,560
-$15,961
$145,000
PCSK9 inhibitor§
85.4
1,254,400
21
2,366,000
$693,450
-$44,627
$274,000
Abbreviations: CV, cardiovascular; ICER, incremental cost-effectiveness ratio; MACE, major adverse cardiovascular event
(nonfatal MI, nonfatal stroke, and cardiovascular death); NNT5, number-needed-to-treat; QALY, quality-adjusted life year.
* In the base case, we assumed that 10% of the population was statin-intolerant. Patients who had a prior history of
cardiovascular disease received incremental treatment with ezetimibe or a PCSK9 inhibitor (n = 1,460,000 in 2015). The analytic
horizon was lifetime (defined as until patients reached the age of 95 years). To reflect the precision of the model, person-years
of treatment are rounded to the 100,000s; MACE and QALYs are rounded to the nearest 100s; costs are rounded to the
millions; and ICERs to the 1000s.
† Number of patients that would need to be treated for 5 years to avert one MACE event.
⌃ All costs are reported in 2015 U.S. dollars. Future costs and QALYs are discounted 3% a year.
§ Both the ezetimibe and PSCK9 inhibitor arms are compared with the control (no additional lipid-lowering therapy) arm.
©Institute for Clinical and Economic Review, 2015
Page ES12
Secondary Prevention Among Patients with a Prior History of CVD and LDL-C ≥ 70mg/dL on Statin
Therapy
Compared with the control arm, treatment with ezetimibe improved outcomes at an ICER of
$135,000/QALY while PCSK9 inhibitors averted 5,621,800 MACE over the lifetime horizon and
produced 10,573,800 additional QALYs at an ICER of $302,000/QALY.
Table ES6. Base-Case Clinical and Economic Outcomes Among Patients with a Prior History of CVD
and LDL-C ≥ 70mg/dL on Statin Therapy.*
Person-years Total MACE
of treatment
averted
(millions)
NNT5†
Statin
QALYs
gained⌃
Incremental
Drug Costs⌃
(million $)
Incremental
Costs, Other CV
Care⌃
(million $)
ICER
($/QALY)
comparator
Statin +
Ezetimibe§
409.1
2,253,800
51
4,345,900
$673,155
-$85,520
$135,000
Statin + PCSK9
inhibitor§
416.9
5,621,800
21
10,573,800
$3,406,692
-$210,702
$302,000
Abbreviations: CV, cardiovascular; ICER, incremental cost-effectiveness ratio; LDL, low-density lipoprotein; MACE, major
adverse cardiovascular event (nonfatal MI, nonfatal stroke, and cardiovascular death); NNT, number-needed-to-treat; PCSK9,
proprotein convertase subtilisin/kexin type 9; QALY, quality-adjusted life year.
* In the base case, patients with pre-existing CVD and LDL-C ≥ 70mg/dL on statin therapy received incremental therapy with
ezetimibe or a PCSK9 inhibitor (n = 7,271,000 in 2015). The analytic horizon was lifetime (defined as until patients reached the
age of 95 years). To reflect the precision of the model, person-years of treatment are rounded to the 100,000s; MACE and
QALYs are rounded to the 100s; costs are rounded to the millions; and ICERs to the 1000s.
† Number of patients that would need to be treated for 5 years to avert one MACE event.
⌃ All costs are reported in 2015 U.S. dollars. Future costs and QALYs are discounted 3% a year.
§ Both the statin+ezetimibe and statin+PSCK9 inhibitor arms are compared with the statin-only arm.
Scenario Analyses
In order to explore possible subpopulations for whom the incremental cost-effectiveness ratios
might be lower, we evaluated the effect of only initiating therapy immediately after an incident MI.
All patients who had an incident, first-ever MI in 2015 who were receiving statin therapy if able to
tolerate it received ezetimibe or a PCSK9 inhibitor. ICERs were lower than in the base case analysis
for all secondary prevention ($170,000/QALY and $74,000/QALY for PCSK9 inhibitors and ezetimibe
respectively) due to a greater reduction in the absolute number of MACE events.
©Institute for Clinical and Economic Review, 2015
Page ES13
Sensitivity Analyses
Across all subpopulations, results were most sensitive to changes in the price of PCSK9 inhibitors
and the length of the time horizon (which was varied from 20 years to the lifetime). However, in
none of the univariate sensitivity analyses except for price did the ICERs for PCSK9 inhibitor therapy
fall below $219,000 per QALY. We varied the effect of PCSK9 inhibitors on cardiovascular event
rates (from -25% to +25% relative to the base case), and found this to be a moderately sensitive
parameter; in the FH population, for example, cost-effectiveness ratios ranged from $250,000 to
$359,000 per QALY gained. Findings from one-way sensitivity analyses are described in further
detail in the full report.
Threshold Analyses
As shown in Table ES7 below, we also evaluated the drug costs at which PCSK9 inhibitors would be
considered cost-effective under conventional willingness-to-pay thresholds of $50,000/QALY,
$100,000/QALY, and $150,000/QALY. Across all subpopulations and thresholds of interest, these
prices represented discounts of 42-78% from the full wholesale acquisition cost of $14,350. When
all patient subpopulations are merged to reflect the entire eligible population, prices were $3,166,
$5,404, and $7,735 to achieve thresholds of $50,000, $100,000, and $150,000 per QALY
respectively.
Table ES7. Threshold analyses: Annual drug cost at which PCSK9 inhibitors would be cost-effective
in subpopulations under varying willingness-to-pay thresholds.*
WTP threshold
Patient Subpopulation
$50,000/QALY
$100,000/QALY
$150,000/QALY
FH on statin (as treated) + statinintolerant †
$3,400
$5,700
$8,000
Pre-existing CVD, LDL-C ≥ 70 mg/dL,
and statin-intolerant ||
$3,400
$5,800
$8,300
Pre-existing CVD, LDL-C≥ 70 mg/dL on
maximally tolerated statin dose ¶
$3,100
$5,300
$7,600
Abbreviations: CVD, cardiovascular disease; FH, familial hypercholesterolemia; LDL, low-density lipoprotein; PCSK9, proprotein
convertase subtilisin/kexin type 9; QALY, quality-adjusted life year; WTP, willingness-to-pay.
* Only drug costs and costs related to cardiovascular care were included in the ICER for these analyses. The analytic horizon was
lifetime (defined as until patients reached 95 years of age), and future costs and QALYs were discounted at 3% a year. To reflect
precision in the model, the reported threshold drug costs are rounded to the nearest 100s.
† Patients who met the operational definition of FH and are either already receiving statin therapy or deemed statin-intolerant
(10% of the population) received incremental therapy with a PCSK9 inhibitor (n = 605,000 in 2015). Complete results of this
analysis are presented in Table 14 in the full report.
|| Ten percent of the population was assumed to be statin-intolerant (n = 1,460,000 in 2015). Complete results of this analysis
are presented in Table 16 in the full report.
¶ Patients with pre-existing CVD and LDL-C ≥ 70mg/dL already receiving statin therapy received incremental therapy with a
PCSK9 inhibitor (n = 7,271,000 in 2015). Complete results of this analysis are presented in Table 17 in the full report.
©Institute for Clinical and Economic Review, 2015
Page ES14
Comparative Value: Health System Value
Budget Impact Model: Methods
We used the same model employed for the care value analysis to estimate total budgetary impact.
Budgetary impact was defined as the total incremental cost of the therapy in each population:
incremental health care costs (including drug costs) minus any offsets in these costs from averted
cardiovascular events. All costs were undiscounted and estimated over one- and five-year time
horizons. The five-year timeframe was of primary interest, given the potential for cost offsets to
accrue from averted cardiovascular events. In addition to FH and patients with a history of CVD who
are (a) statin intolerant or (b) not at LDL-C target on statin therapy, we also considered the
budgetary impact if the treated population were limited to the higher-risk subset of patients with a
history of CVD who received PCSK9 inhibitors immediately following an incident (i.e., first-ever) MI
in 2015. Our calculations assume that utilization of new drugs is “unmanaged” – i.e., without payer
or pharmacy benefit management controls in place – to provide an upper bound for likely patterns
of drug uptake by five years after launch.
We examine six characteristics of the drug and marketplace to estimate unmanaged drug uptake.
These characteristics are listed below:

Magnitude of improvement in clinical safety and/or effectiveness

Patient-level burden of illness

Patient preference (ease of administration)

Proportion of eligible patients currently being treated

Primary care vs. specialty clinician prescribing/use

Presence or emergence of competing treatments of equal or superior effectiveness
Based on our assessment of these criteria, we assign a new drug to one of four categories of
unmanaged drug uptake patterns: 1) very high (75% uptake by year 5); 2) high (50% uptake by year
5); 3) intermediate (25% uptake by year 5); and 4) low (10% uptake by year 5). We then compare
our estimates to a budget impact threshold that represents a potential trigger for policy
mechanisms to improve affordability through changes to pricing, payment, or patient eligibility. As
described in ICER’s methods presentation (http://www.icer-review.org/wpcontent/uploads/2014/01/Value-Assessment-Framework-9-7.pdf), this threshold is based on an
underlying assumption that health care costs should not grow much faster than growth in the
overall national economy. From this foundational assumption our potential budget impact
threshold is derived using an estimate of growth in US gross domestic product (GDP) +1%, the
average number of new molecular entity approvals by the FDA each year, and the contribution of
spending on retail and facility-based drugs to total health care spending. Therefore, according to
©Institute for Clinical and Economic Review, 2015
Page ES15
our calculations, for 2015-16, the five-year annualized potential budget impact threshold that
should trigger policy actions to manage affordability is calculated to total approximately $904
million per year. In this report, each PCSK9 inhibitor is considered as an individual new drug, so the
budget impact threshold for each drug is $904 million, and $1.8 billion for the two drugs combined.
We combine consideration of the potential budget impact with the prices necessary to meet
commonly accepted societal willingness-to-pay thresholds in order to calculate a value-based price
benchmark for each new drug. This price benchmark begins with the “care value” price range
needed to achieve cost-effectiveness ratios of $100,000-$150,000 per QALY for the population
being considered, but the value-based price benchmark has an upper limit determined by the price
at which the new drug would exceed the potential budget impact threshold of $904 million. If the
potential budget impact does not exceed $904 million, then the value-based price benchmark
remains the full care value price range.
Results
Results from the budget impact model showed that if both the FH and CVD populations were
treated with the uptake pattern assumptions mentioned above, 527,000 individuals would receive
PCSK9 therapy in the first year. After one year of PCSK9 treatment, cost offsets due to reduced
cardiovascular adverse events ranged from $592 per patient with FH to $1,010 per patient for
patients with CVD who are statin-intolerant. Including this cost offset, one-year budget impact is
still estimated to be quite high: approximately $7.2 billion for all patient populations.
As uptake of new PCSK9 inhibitors is estimated to increase over the entire 5-year time horizon, we
estimate that approximately 2.6 million persons would receive PCSK9 inhibitor therapy for one or
more years by the end of that period. Total budgetary impact over five years is estimated at
approximately $19 billion, $15 billion, and $74 billion for the FH, CVD statin-intolerant, and CVD not
at LDL-C target subpopulations, respectively. When these 5-year budget impact figures are
annualized, they equal $21.4 billion in net health care cost growth per year, which is well above the
budget impact threshold of $1.8 billion for the two drugs combined. In order to not exceed this
budget impact threshold, approximately 1% of eligible patients could be treated at the average list
price of $14,350 per year.
Figure ES1 on the following page provides findings of multiple analyses that give perspective on the
relationship between varying possible drug prices, cost-effectiveness ratios, drug uptake patterns,
and potential budget impact.
As can be seen in Figure ES1, even at a drug cost of $3,166 dollars per year, the cost at which the
cost/QALY = $50,000, if 50% of all eligible patients are ultimately treated over a five-year time
period the annualized budget impact is approximately $5.6 billion per year. At the list price of
$14,350 used for this report, if only 25% of eligible patients receive treatment, the annualized
©Institute for Clinical and Economic Review, 2015
Page ES16
budget impact is nearly $19 billion, meaning that over the five-year period a total of almost $100
billion would have been added to health care costs in the United States.
Figure ES1. ICER value graph combining cost-effectiveness and potential budget impact analyses.
Colored lines represent the impact on annualized budget impact of different uptake patterns
(eligible patients treated) at the actual list price of the drug (dashed line) and at drug prices needed
to achieve common incremental cost-effectiveness ratios.
Estimated Cost/QALY: $296,850
($14,350 annual drug price)
$36
Annual Budget Impact (Billions)
$32
Cost/QALY $150,000
($7,735 annual drug
price)
$28
Cost/QALY $100,000
($5,404 annual drug
price)
$24
$20
$16
Cost/QALY $50,000
($3,166 annual drug
price)
$12
$8
$4
$0
0%
25%
50%
75%
100%
% Eligible Patients Treated
Draft Value-based Benchmark Prices
Our draft value-based benchmark prices for each key subpopulation and for the overall treated
population are provided in Table ES8 on the following page. Detailed calculations for the valuebased price benchmarks presented below are available in Appendix Table 15.
As shown in the table on the following page, if only the FH or the CVD statin-intolerant populations
were treated, the entire care value price range is lower than the price at which the potential budget
impact threshold would be exceeded. Thus, the value-based price benchmark for these two
subpopulations is the care value price range. This is not surprising given the relatively small size of
each of these populations. In contrast, the care value price range for the much larger population of
patients with CVD not at LDL-C target is higher than the maximum price that would not exceed the
budget impact threshold.
©Institute for Clinical and Economic Review, 2015
Page ES17
When all subpopulations are combined, the care value price range is $5,404-$7,735. However, this
price range is higher than the maximum price that could be charged before exceeding the potential
budget impact threshold ($2,177). Therefore, the draft ICER value-based price benchmark for each
of the new PCSK9 inhibitor drugs, with all the assumptions mentioned previously regarding 5-year
uptake patterns and cost offsets, is $2,177. This figure represents an 85% discount from the full
wholesale acquisition cost assumed in our analysis ($14,350).
Table ES8. Draft value-based price benchmarks for PCSK9 inhibitor therapy.
Population
Care Value Price:
$100K/QALY
Care Value Price:
$150K/QALY
Max Price at
Potential Budget
Impact Threshold
Draft Value-Based
Price Benchmark
FH
(n=453,443)
CVD statin-intolerant
(n=364,948)
CVD not at LDL target
(n=1,817,788)
$5,700
$8,000
$10,278
$5,700-$8,000
$5,800
$8,300
$12,896
$5,800-$8,300
$5,300
$7,600
$2,976
$2,976
TOTAL (n=2,636,179)
$5,404
$7,735
$2,177
$2,177
FH: familial hypercholesterolemia; CVD: cardiovascular disease; LDL: low-density lipoprotein; QALY: qualityadjusted life year
Summary and Comment
The results of our cost-effectiveness analysis suggest that the use of PCSK9 inhibitors may produce
substantial reductions in non-fatal MIs, non-fatal strokes, and cardiovascular deaths over the
lifetime analytic horizon. The NNT5 (number of patients that would be needed to be treated for 5
years to avoid one major adverse cardiovascular event) of 28 for PCSK9 inhibitors appears to be
relatively low; despite this, treatment with PCSK9 inhibitors generates cost-effectiveness ratios that
far exceed commonly-accepted thresholds, such as $100,000/QALY.79 Achieving cost-effectiveness
at a threshold of $100,000/QALY would require price reductions of 60% to 63% compared with
current prices. And the results of our analysis of potential budget impact suggest that even deeper
reductions may be required to avoid excessive cost burdens to the health care system. Our valuebased price benchmark for each PCSK9 inhibitor is $2,177 annually, which represents an 85%
reduction from the list price of $14,350.
©Institute for Clinical and Economic Review, 2015
Page ES18
1. Background
1.1 Introduction
The focus for this assessment is the use of alirocumab and evolocumab for individuals with elevated
LDL cholesterol. We assess the evidence on the comparative effectiveness and value of the drugs
across relevant populations including:

Patients with familial hypercholesterolemia

Patients with established cardiovascular disease

Patients at elevated risk for cardiovascular disease
Scope of the Assessment:
The scope for this assessment is described on the following page using the PICOTS (Population,
Intervention, Comparators, Outcomes, Timing, Settings) framework. The evidence review is based
on the 25 clinical trials and two published systematic reviews and meta-analyses. The results were
cross-checked with the manufacturers’ FDA submission documents and the FDA’s briefing
documents.
Figure 1: Analytic Framework for the Assessment
Health Care Utilization Outcomes
Treatment
PCSK9 inhibitor
Lower LDL-C
•
•
Hospitalization
Days alive outside hospital
Individuals
with elevated
LDL-C
Clinical and Patient-Centered
Outcomes
Subgroups:
•
Familial
hypercholesterolemia
•
Known CVD
•
High risk for CVD events
•
Statin intolerance
•
•
•
•
•
•
©Institute for Clinical and Economic Review, 2015
Mortality
CVD Mortality
Non-fatal MI
Non-fatal stroke
Unstable angina requiring
hospitalization
Adverse events
Page 1
Population
The populations of interest include:

Individuals with heterozygous familial hypercholesterolemia (HeFH) OR homozygous
familial hypercholesterolemia (HoFH) whose cholesterol levels are not at goal

Individuals with known cardiovascular disease (CVD) who are intolerant of statins or
whose cholesterol levels are not at goal

Individuals who are at high risk for CVD who are intolerant of statins or whose
cholesterol levels are not at goal
Interventions
The interventions are the following PCSK9 inhibitors considered as a class:


Alirocumab (Praluent®, Sanofi and Regeneron Pharmaceuticals, Inc.)
Evolocumab (Repatha™, Amgen)
We considered the PCSK9 inhibitors as a class rather than separately for several reasons. First, there
are no randomized trials comparing the two, which would allow for direct comparison of the LDLlowering effects. Second, network meta-analytic techniques are not yet available to perform
indirect comparisons for continuous outcomes such as the percentage reduction in LDL-C. Third, the
magnitude of the reduction in LDL-C with PCSK9 inhibitors is much greater than any potential
differences between the different drugs or their dosing. Finally, the number of clinical events for
the individual PCSK9 inhibitors is too small to offer meaningful comparisons.
Comparators
The studies compare the PCSK9 inhibitors to usual care (i.e., statin therapy, lifestyle and dietary
changes), placebo, and/or to ezetimibe.
Outcomes
Outcomes of interest include the impact of cholesterol-lowering interventions on:






Mortality
CVD mortality
CVD events (myocardial infarction, stroke, unstable angina, revascularization)
LDL-C reduction as an intermediate marker
Short- and long-term complications and adverse events including neurocognitive events,
myalgias, and local injection site reactions
Economic outcomes, including payer costs, patient productivity, and cost-effectiveness
©Institute for Clinical and Economic Review, 2015
Page 2
Timing
Evidence on intervention effectiveness was limited to phase 2 or 3 comparative studies with at least
two months of follow-up for LDL-C reduction. Evidence on cardiovascular outcomes and harms was
derived from comparative studies of any duration.
Settings
All relevant settings were considered, including inpatient, clinic, and outpatient settings.
©Institute for Clinical and Economic Review, 2015
Page 3
2. The Topic in Context
Cardiovascular disease (CVD)
Cardiovascular disease is the most common cause of death in the United States and approximately
one third of American adults have CVD.3 The American Heart Association (AHA) defines
cardiovascular disease as anyone with a history of myocardial infarction (MI), stroke, angina,
congestive heart failure (CHF), peripheral artery disease, or hypertension. Major cardiovascular
events in clinical trials usually include death due to cardiovascular disease, non-fatal MI, non-fatal
stroke, unstable angina requiring hospitalization, and revascularization (stenting, bypass surgery).
Low density lipoprotein cholesterol (LDL) “lower is better” hypothesis
Low density lipoprotein cholesterol (LDL-C) is a major modifiable risk factor for myocardial
infarction, stroke, and death from cardiovascular disease.3,80 Multiple randomized clinical trials have
demonstrated that lowering LDL-C with statin therapy reduces the risk of myocardial infarction,
stroke, and death from cardiovascular disease.81-83 Many investigators believe that the greater the
reduction in LDL-C the greater the reduction in cardiovascular events, but the topic remains
controversial.84-88 However, several drugs that lower LDL-C – including hormone therapy, niacin, and
torcetrapib – have not decreased cardiovascular disease events when evaluated in randomized
trials despite lowering LDL-C.89-93 Torcetrapib lowered LDL-C by 25%, but in the pivotal 15,000
person randomized trial, torcetrapib increased cardiovascular events by 25% and total mortality by
58%.90 On the other hand, the recently published IMPROVE-IT trial demonstrated that the lowering
of LDL-C with ezetimibe significantly reduced cardiovascular event rates by 6% (95% CI 1 to 11%)
after a median follow-up of approximately 5 years.78
Guidelines for cholesterol lowering therapy
In 2013, the ACC/AHA released updated guidelines for the treatment of cholesterol in order to
reduce cardiovascular risk in adults.4 Diet therapy is recommended for all patients. The guidelines
make strong recommendations for high intensity statin therapy to treat individuals with
cardiovascular disease who are ≤ 75 years of age; moderate intensity statin use in individuals with
diabetes mellitus and LDL-C levels between 70 and 189 mg/dL who are ages 40-75 years of age; and
for high intensity statin use in individuals with a 10-year risk for cardiovascular disease ≥ 7.5% and
LDL-C levels between 70 and 189 mg/dL who are ages 40-75 years of age. The guidelines make
moderate recommendations for high intensity statin therapy to treat individuals with LDL-C levels ≥
190 mg/dL who are ≥ 21 years of age.
Statin therapy is the primary therapy indicated for the treatment of high LDL-C. High intensity statin
therapy includes atorvastatin 40 - 80 mg daily and rosuvastatin 20 - 40 mg daily. Moderate intensity
statin therapy includes atorvastatin 10 - 20 mg daily, rosuvastatin 5 - 10 mg daily, simvastatin 20-40
©Institute for Clinical and Economic Review, 2015
Page 4
mg daily, pravastatin 40-80 mg daily, lovastatin 40 mg daily, fluvastatin XL 80 mg daily, fluvastatin
40 mg twice daily, and pitastatin 2-4 mg daily.
The major change in the 2013 ACC/AHA guidelines compared to the earlier National Cholesterol
Education Program Adult Treatment Panel III (NCEP ATP III) guideline5 was moving away from
recommending specific LDL-C levels as treatment targets. In the prior guidelines, statin therapy was
recommended to reach a target LDL-C level of < 100 mg/dL for individuals with cardiovascular
disease and those with a 10-year risk ≥ 20%. For individuals with multiple risk factors and a 10-year
risk < 20%, the target LDL-C level was < 130 mg/dL. The 2011 European guidelines also recommend
statin therapy to reach a target LDL-C level of < 70 mg/dL for individuals with cardiovascular disease
or diabetes and < 100 mg/dL for primary prevention in high risk individuals.6 A more complete
discussion of all guidelines, as well as those from other organizations, can be found in Appendix A2.
Need for additional therapy
The use of statins to decrease LDL-C has contributed to the marked decline in death from CVD since
1950, but some patients are not able to tolerate statins and others have inadequate reductions in
LDL-C.94 Patients with familial hypercholesterolemia are the largest group of patients who may have
inadequate reductions in LDL-C with statins due to their high baseline levels of LDL-C.
Familial hypercholesterolemia (FH)
Familial hypercholesterolemia (FH) is an autosomal dominant inherited condition that causes
elevated LDL-C in both the heterozygous (HeFH) and homozygous (HoFH) states.9-12 Individuals with
HoFH have LDL-C levels > 500 mg/dL and experience cardiovascular events by age 20. It is an
extremely rare condition (~1 case per 1million people) with only 300-400 individuals in the US
affected by HoFH. HeFH is more common, with estimates of affected individuals in the US varying
between 500,000 to over a million, though there is considerable uncertainty in this estimate.9-12
Individuals with HeFH often have LDL-C levels that are approximately two to three times normal
(i.e., 250-350 mg/dL).13,14
FH is usually diagnosed on the basis of clinical criteria because it is caused by mutations in several
different genes, not all of which have been identified. The diagnostic criteria include a family history
of early onset CVD, an elevated LDL-C level (>190 mg/dL in adults and >160 mg/dL in children),
physical exam findings of tendon xanthomata (cholesterol deposits in tendons) or corneal arcus
before the age of 45 years, and DNA analysis for known deleterious mutations causing FH. 9-12
Treatment is usually initiated early with high intensity statin therapy, with the addition of LDL
apheresis in those patients with an inadequate response to aggressive lipid lowering therapy.
©Institute for Clinical and Economic Review, 2015
Page 5
Statin intolerance
Statin intolerance is primarily due to muscle symptoms. These range from asymptomatic mild
elevations in creatinine kinase (CK), a muscle enzyme, to muscle aches (myalgias) with or without
mild elevations in CK (<4 times the upper limit of normal), to frank myositis (CK ≥ 4 time the upper
limit of normal).15,16 The muscle symptoms can include weakness, pain, stiffness and cramps. In the
randomized trials of statins, the incidence of muscle symptoms was less than 5%, but these trials
often excluded patients who did not tolerate statins during a run in period prior to randomization
and the participants are not representative of the general population. Two studies specifically
examined statin intolerance in clinical practice. The Prediction of Muscular Risk in Observational
(PRIMO) trial reported a 10% incidence of mild to moderate muscle symptoms for patients on high
intensity statin therapy.17 Similarly, the Effect of Statins on Skeletal Muscle Function and
Performance (STOMP) study reported a 9.4% incidence of muscle symptoms in statin-naïve patients
treated with atorvastatin 80 mg daily compared to a 4.6% incidence in patients randomized to
placebo.18 Risk factors for muscle symptoms include older age, female sex, and Asian race.
Strategies to manage muscle symptoms include reducing the dose of the statin, switching to a
different statin, and treating vitamin D deficiency.16,95
The two trials of evolocumab that specifically enrolled statin intolerant patients (GAUSS, GAUSS 2)
used slightly differing definitions.61,62 For the GAUSS trial, participants were required to fail at least
one statin due to intolerable muscle symptoms on the lowest dose of the statin.62 For the GAUSS 2
trial, participants were required to fail at least two statins due to intolerable muscle symptoms on
the lowest dose of the statin.61
Precise measurement of statin intolerance is difficult because muscle symptoms arising from other
causes are common, particularly in older individuals. As an example, during the run-in period of the
ODYSSEY ALTERNATIVE trial, the randomized trial of alirocumab in statin-intolerant patients, 49% of
patients were reported to be intolerant of placebo due to musculoskeletal complaints.43,44
Furthermore, 70% of the patients randomized to receive atorvastatin 20 mg daily (blinded) in the
same trial tolerated it for the 24 week duration of the trial. Thus, many patients labeled as statinintolerant may actually be able to tolerate a statin.
Ezetimibe (Zetia®)
Ezetimibe is a drug that inhibits the absorption of cholesterol in the intestines. The FDA approved
ezetimibe 10 mg daily in October 2002 for LDL-C-lowering alone or in conjunction with other lipid
lowering therapies in patients with hypercholesterolemia including homozygous familial
hypercholesterolemia. Meta-analyses of randomized trials report that ezetimibe 10 mg lowers LDLC by 23.6% (95 %CI 21.7 to 25.6%) when added to statin therapy19 and by 18.6% (95% CI 17.5 to
19.7%) as monotherapy.20 However, treatment with ezetimibe has been controversial because of
negative findings in the ARBITER-6 and ENHANCE trials.21,22 These were small trials that were not
©Institute for Clinical and Economic Review, 2015
Page 6
designed to evaluate cardiovascular outcomes. The IMPROVE-IT trial randomized 18,144 patients
hospitalized for an acute coronary syndrome in the prior 10 days to the combination of simvastatin
and ezetimibe or simvastatin and placebo and followed them for a median of approximately 5
years. The hazard ratio for the reduction of all cardiovascular events was 0.936 (95% CI 0.89 to
0.99). The estimated cumulative event rate at 7 years was 32.7% in the ezetimibe group and 34.7%
in the placebo group (p=0.016). The publication of the IMPROVE-IT trial in 2015 has renewed
enthusiasm for the use of ezetimibe to lower LDL-C beyond the reduction achieved with statin
therapy, even though the relative and absolute benefits were small.
Proprotein convertase subtilisin/kexin type 9 (PCSK9) and cardiovascular disease
PCSK9 is a protein found in the circulation that can bind to LDL-receptors. When the two undergo
endocytosis, the LDL receptor is broken down and not recycled to the surface.96 Thus higher levels
of PCSK9 reduce the number of LDL receptors. If there are fewer LDL receptors, then LDL
cholesterol levels rise in the blood. Conversely, lower levels of PCSK9 in the blood leads higher LDL
receptor density and lower levels of LDL-C in the blood.
In 2003, a gain of function mutation was found in the PCSK9 gene that increases its activity, lowers
LDL-receptor density, and causes high levels of LDL-C.96-99 Patients with this mutation are at
increased risk for premature cardiovascular disease.96,100,101 Subsequently, loss of function
mutations were identified that decrease the activity of PCSK9 and cause low levels of LDL-C.102-104
Patients with these mutations are at decreased risk for cardiovascular disease.103,105
The biology described above suggests that drugs targeting PCSK9 have the potential to reduce LDL
and cardiovascular disease. In June 2015, the FDA advisory panel voted to recommend approval of
two human monoclonal antibodies that target proprotein convertase subtilisin/kexin type 9 (PCSK9)
in the blood and markedly reduce LDL-C levels.
Alirocumab (Praluent®, Sanofi and Regeneron Pharmaceuticals, Inc.)
Alirocumab is a human monoclonal antibody that inhibits PCSK9. It is administered as a
subcutaneous injection once every two weeks at doses of either 75 mg or 150 mg. It can be given as
primary therapy to lower LDL-C, or it can be used in combination with statin therapy. Combination
therapy is particularly efficacious, as statin therapy has been shown to up-regulate the production
of PCSK9. The FDA approved alirocumab in July 2015; approved indications for alirocumab include
use in addition to diet and maximally tolerated statin therapy in adult patients with (a)
heterozygous familial hypercholesterolemia (HeFH); or (b) patients with clinical atherosclerotic
cardiovascular disease such as heart attacks or strokes, who require additional lowering of LDL
cholesterol. The currently-listed wholesale acquisition cost of alirocumab is $14,600 annually, which
is nearly 60 times the cost of generic statins.
©Institute for Clinical and Economic Review, 2015
Page 7
Evolocumab (Repatha™, Amgen)
Evolocumab is a human monoclonal antibody that inhibits PCSK9. It is administered as a
subcutaneous injection 140 mg once every two weeks or 420 mg once every four weeks. As noted
for alirocumab, combination therapy with statins is particularly efficacious, as statin therapy has
been shown to up-regulate the production of PCSK9. The FDA approved evolocumab in August
2015; approved indications for evolocumab include use in addition to diet and maximally tolerated
statin therapy in adult patients with (a) heterozygous familial hypercholesterolemia (HeFH); (b)
homozygous familial hypercholesterolemia (HoFH); or (c) patients with clinical atherosclerotic
cardiovascular disease such as heart attacks or strokes, who require additional lowering of LDL-C.
The currently-listed wholesale acquisition cost of evolocumab is $14,100 annually.
©Institute for Clinical and Economic Review, 2015
Page 8
3. Summary of Coverage Policies
Due to their recent approval, coverage policies and authorization criteria for PCSK9 inhibitors are
emerging. Below is a summary of available coverage policies as of October 2015. For completeness,
we also document coverage policies in Table 1 that pertain to ezetimibe alone or in combination
with simvastatin (Vytorin®) as well as rosuvastatin (Crestor®), the lone high-intensity statin available
in branded form. Further explanation of these policies is available in Appendix 3.
3.1 Summary of Coverage for PCSK9 inhibitors: Praluent and Repatha
Regional Private Payers
Blue Cross Blue Shield of Vermont
http://www.bcbsvt.com/wps/wcm/connect/ae48547e-d7d9-444a-8d8b-d060db8f50fb/2015-praluentpa-guidelines.pdf?MOD=AJPERES; http://www.bcbsvt.com/wps/wcm/connect/975504d0-cca6-41f78943-6c10577b8d86/2015-repatha-pa-guidelines.pdf?MOD=AJPERES
Coverage of both Praluent and Repatha is subject to prior authorization criteria. To receive a
prescription, patients must have a diagnosis of HoFH, or HeFH with failure to reach LDL-C goals after
trying at least one high intensity statin for 60 days. Praluent may also be covered as secondary
prevention for atherosclerotic cardiovascular disease (ASCVD) or primary prevention for diabetes in
patients who have not met LDL-C goals after trials of two high intensity statins for at least 60 days,
or in patients who have experienced adverse effects with trials of at least two statins. Patients must
be on a low fat diet and must be at least 18 years of age. A cardiologist must issue the initial request
for either medication. Initial approvals will be effective for 6 months, and can be renewed if there
is evidence of LDL-C reduction. Renewals are effective for 24 months and may be prescribed by
another physician in consultation with a cardiologist. Dosing for Praluent should start at 75mg every
two weeks subcutaneously. For HeFH, dosing for Repatha should start at 140mg once every 2
weeks, or 420mg once monthly. For HoFH, dosing should start at 420mg once monthly.
ConnectiCare
http://www.connecticare.com/globalfiles/pharmacycentral/ConnectiCare%20Formulary%20%20Chart.pdf
Coverage of Praluent is subject to prior authorization criteria. Prescriptions must be filled by a
specialty pharmacy. Repatha is not publicly listed in ConnectiCare’s formulary at the time of this
publication.
©Institute for Clinical and Economic Review, 2015
Page 9
Harvard Pilgrim Health Care
https://www.harvardpilgrim.org/portal/page?_pageid=213,57031&_dad=portal&_schema=PORTAL
Harvard Pilgrim Health Care negotiated a discounted payment rate as well as the first pay-forperformance contract for Amgen’s PCSK9 inhibitor Repatha. The drug will be included on HPHC’s
formulary at a reduced price. In addition, Amgen will provide additional rebates if the drug doesn’t
reduce cholesterol to specified target levels for eligible patients. Amgen may also pay out rebates if
more patients covered by HPHC’s plans end up taking the drug than was initially projected. The
exact terms of the agreement including the level of discount and rebate amounts are not publicly
available at this time.
National Payers
Aetna
http://www.aetna.com/products/rxnonmedicare/data/2016/MISC/PCSK9.html
Coverage of both Praluent and Repatha is subject to precertification criteria. Patients must have a
documented diagnosis of HeFH or existing cardiovascular disease. In addition, patients must have
LDL-C >70mg/dl after trying at least 2 different treatment regimens, including a high-potency statin
at the maximally tolerated dose in combination with ezetimibe. Patients must have tried each
regimen for at least 4 weeks with optimal compliance. Praluent and Repatha must be used in
combination with a statin at maximally tolerated dose. Patients must be at least 18 years of age,
have triglyceride levels <400mg/dl, have no history of severe renal impairment. Female patients
must not be pregnant or planning to become pregnant while using either drug.
Repatha may also be covered for patients with HoFH who are at least 13 years of age. In these
cases, Repatha must be used in combination with lipid-lowering therapy including a statin,
ezetimibe, or lipid apheresis.
Anthem
https://www.anthem.com/ca/medicalpolicies/policies/mp_pw_c182635.htm
Praluent and Repatha are covered for patients who are at least 18 years old and at high risk for
Acute Coronary Syndrome (ACS). Risk is identified by presence of HoFH, HeFH, or a history of
ASCVD. To be eligible, patients with these conditions must be on high intensity statin therapy, have
a condition that is a contraindication for statin therapy, or have a statin intolerance. Intolerance is
defined as inability to tolerate at least 2 statin regimens, at least one of which was prescribed at the
lowest starting daily dose; continued symptoms despite an attempt at dose reduction instead of
©Institute for Clinical and Economic Review, 2015
Page 10
discontinuation; resolution of symptoms with discontinuation of statin therapy; and a return of
symptoms after re-starting statin therapy in patients for whom re-challenge is clinically appropriate.
Other possible causes of symptoms, such as hypothyroidism, drug interactions, concurrent illness,
significant changes in physical activity, or underlying muscle disease, must be ruled out.
Patients meeting these criteria must also be taking ezetimibe in addition to statin therapy (for
patients able to tolerate statins) and have had less than a 50% reduction in LDL-C after at least 90
days of compliant use of lipid lowering therapy and lifestyle modifications. Individuals whose initial
LDL-C is unknown must have documented cardiovascular disease and LDL-C > 70m/dL, or no
documented cardiovascular disease and LDL-C > 100mg/dL.
For continuation of therapy with a PCSK9 inhibitor after initial approval, all criteria must be met and
documentation of LDL-C reduction must be provided.
In addition to the above uses, Repatha is also covered for patients 13 and older with HoFH,
confirmed by presence of two mutant alleles, or confirmed by an untreated LDL-C of >500mg/dl or
treated LDL-C >300mg/dl, with the presence of cutaneous or tendinous xanthoma before age 10, or
untreated LDL-C levels consistent with HeFH in both parents.
United Healthcare
https://www.unitedhealthcareonline.com/ccmcontent/ProviderII/UHC/enUS/Assets/ProviderStaticFiles/ProviderStaticFilesPdf/Tools%20and%20Resources/Policies%20and%
20Protocols/Medical%20Policies/Ox_MPUB_Future_Pharmacy/Med_Nec_Praluent.PDF
Both Praluent and Repatha are approved based on submission of medical records showing HeFH
confirmed by pre-treatment LDL-C >190mg/dl (or >155mg/dL in patients under the age of 16) in
both patient and in adult first- or second-degree relative. They are also covered for patients with
ASCVD. Medical records must document that the patient has received at least 12 weeks or highintensity statin therapy and will continue to receive a high-intensity statin. In patients unable to
tolerate high-intensity statin due to documented myalgia or myositis, a moderate- or low-intensity
statin may be used. If the patient is unable to tolerate all doses of statin therapy, and the patient
has undergone a statin re-challenge with a different low-intensity statin with a documented return
of muscle pain, has a contraindication to statin use, or has experienced rhabdomyolysis with statin
treatment with CK elevations <10 times ULN, a PSCK9 inhibitor may be covered. Patients must also
have tried Zetia in combination with statin therapy, or have a contraindication to Zetia. The drugs
are also covered for patients with LDL-C >100mg/dL with ASCVD while on maximally tolerated lipid
lowering therapy, or LDL-C >130mg/dL without ASCVD. The drugs must be used in addition to a lowfat diet and exercise and must be prescribed by a cardiologist, endocrinologist, or lipid specialist.
They cannot be used in combination with any other PCSK9 inhibitor.
©Institute for Clinical and Economic Review, 2015
Page 11
Initial authorization is valid for 6 months, after which point the prescription may be renewed. To
meet criteria for renewal, patients must continue to received maximally tolerated statin therapy,
continue to receive Zetia in addition to statin therapy, continue a low-fat diet and exercise program,
and be prescribed by one of the aforementioned specialists. Medical records indicating an LDL-C
reduction must be submitted. Re-authorization is valid for 12 months.
Repatha is also covered for HoFH confirmed by medical records documenting a pre-treatment LDL-C
>500mg/dL or a treated LDL-C >300mg/dL and presence of either xanthoma before age 10 or
evidence of HeFH in both parents. Patients should be on a low-fat diet and exercise program and be
on other lipid-lowering therapies. Prescriptions must come from a specialist as described above.
Repatha should not be used in combination with Juxtapid or Kynamro. The same re-authorization
criteria as above apply.
3.2 Summary of Coverage for Existing Lipid-Lowering Therapies
Table 1 on the following page summarizes coverage policies for other key lipid-lowering therapies.
As displayed in the table, nearly all regional public payers as well as regional and national private
payers impose coverage restrictions on Crestor®, Vytorin®, and Zetia®. Detailed descriptions of
these coverage policies can be found in Appendix 3.
©Institute for Clinical and Economic Review, 2015
Page 12
Table 1: Coverage Policies for Crestor, Vytorin, and Zetia
Public Payers
Connecticut
Maine
Crestor®
Vytorin®
Zetia®
Covered
Covered
-Covered
Covered for patients with inadequate response to
atorvastatin dose of at least 80mg/day (or another
equipotent statin), or adverse reaction or
contraindication to atorvastatin
Non-preferred
Covered for patients with inadequate response to
atorvastatin dose of at least 80mg/day (or another
equipotent statin), or adverse reaction or
contraindication to atorvastatin
Non-preferred
Rhode Island
PA and ST required
Vermont
QL apply.
Regional Private Payers
BCBS MA
Tier 2, QL apply and ST required
BCBS RI
Tier 2
BCBS VT
Covered, no restrictions listed
ConnectiCare
Tier 2; ST required and QL apply
HPHC
Tier 2
Deductible exemption through Preventative Drug
Benefit
NHPRI
Tier 3, ST required
THP
Tier 3, PA required
National Private Payers
Aetna
Tier 2, QL
Anthem
Covered
Cigna
Tier 2, PA and ST required for 5mg and 10mg; 30mg
and 40mg covered without restriction
Deductible exemption through Preventative Drug
Benefit
PA required.
PA required.
-Covered. PA required as add-on to Lipitor. No PA
for statin intolerance or patients at maximally
tolerated statin dose.
PA required. Covered for patients with
inadequate response to atorvastatin 80mg/day
or another statin with equipotent dosing, or for
statin-intolerant patients
Non-preferred agent. Must fail with 2 highpotency statins and combination products.
PA required.
PA required.
ST required and QL apply
Tier 3
-Tier 3, ST required and QL apply
Tier 3, QL; ST required for 10/10mg or 10/20mg
formulations. Deductible exemption through
Preventative Drug Benefit
Tier 3, PA required
Tier 2
Tier 3, ST required.
Tier 2
-Tier 2, QL apply
Tier 2, ST required.
Deductible exemption through Preventative Drug
Benefit
Tier 3, PA required
Tier 3
Tier 3, ST required and QL apply
-Tier 3, PA and ST required
Tier 2, QL apply
-Tier 2
Deductible exemption through Preventative Drug
Benefit
Humana
UHC
Tier 2 or 3 depending on plan, QL apply
Tier 3 or 4 depending on plan, QL apply
Tier 2, QL apply
Tier 3 or 4 depending on plan, QL apply
Massachusetts
New Hampshire
Tier 2, QL
Tier 2, QL
QL=Quantity Limits
ST=Step Therapy PA=Prior Authorization --- = Not listed in formulary
©Institute for Clinical and Economic Review, 2015
Page 13
4. Comparative Clinical Effectiveness
4.1 Methods
The goal of this report is to evaluate the comparative clinical effectiveness and comparative value of
PCSK9 inhibitors as a class for patients with elevated LD-C. We have attempted to identify all
randomized controlled trials that evaluated the safety and efficacy of the two FDA approved PCSK9
inhibitors alirocumab and evolocumab. The published meta-analyses found that the LDL-C lowering
effect of the two PCSK9 inhibitors were similar and there are no head to head trials that compare
alirocumab to evolocumab; accordingly, we have elected to examine the impact of PCSK9 inhibitors
as a class.
We searched the Medline database, Embase, Cochrane clinical trials database, Cochrane reviews
database, and the Database of Abstracts of Reviews of Effects (DARE), using the key words
“alirocumab” OR “evolocumab” OR “PCSK9 antibody.” The search was performed for the period
from 1945 through August 7, 2015. No language restriction was used. Full details of the search are
in Appendix A1. The bibliographies of systematic reviews and key articles were manually searched
for additional references. The abstracts of citations were reviewed for relevance and all potentially
relevant articles were reviewed in full.
We included all phase 2 and 3 randomized trials evaluating either alirocumab or evolocumab that
reported adverse events, LDL-C outcomes, or cardiovascular events. We excluded animal studies
and phase 1 studies.
We abstracted data from each trial on the number of patients randomized, the duration of followup, age, sex, diabetes, heart disease, lipid levels, lipid therapy, trial quality measures, and the
experimental and control interventions. We extracted data for intervention groups that evaluated
the FDA approved doses for alirocumab and evolocumab. Key outcomes included changes in LDLcholesterol levels, cardiovascular events, liver and muscle enzyme changes, neurocognitive
outcomes, total adverse events, serious adverse events, discontinuations due to adverse events,
and common adverse events.
The quality of individual studies was assessed by considering the domains listed below, which are
adapted from the methods guide of the Agency for Healthcare Research & Quality (AHRQ 106):
•
•
•
•
•
•
•
Similarity of baseline characteristics and prognostic factors between comparison groups
Well-described methods for randomization and concealment of treatment assignment
Use of valid, well-described primary outcomes
Blinding of subjects, providers, and outcome assessors
Intent-to-treat analysis (all randomized subjects included)
Limited and non-differential loss to follow-up
Disclosure of any conflicts of interest
©Institute for Clinical and Economic Review, 2015
Page 14
We also adopted the approach of the ICER Evidence Rating Matrix (see Figure 2 on the following
page) to evaluate the overall strength of evidence for each therapy (ICER Evidence Rating Matrix:
http://www.icer-review.org/wp-content/uploads/2008/03/Rating-Matrix-User-Guide-FINAL-v1022-13.pdf)
The evidence rating reflects a joint judgment of two critical components:
1. The magnitude of the difference between a therapeutic agent and its comparator in “net
health benefit” – the balance between clinical benefits and risks and/or adverse effects AND
2. The level of certainty in the best point estimate of net health benefit.
Figure 2. ICER Evidence Rating Matrix
Comparative Clinical Effectiveness
High
Certainty
D
C
B
A
B+
Moderate
Certainty
C+
P/I
I
Low
Certainty
I
I
Negative
Net Benefit
I
Comparable
Net Benefit
Small
Net Benefit
Substantial
Net Benefit
A = “Superior” - High certainty of a substantial (moderate-large) net health benefit
B = “Incremental” - High certainty of a small net health benefit
C = “Comparable”- High certainty of a comparable net health benefit
D=”Negative”- High certainty of an inferior net health benefit
B+=”Incremental or Better” – Moderate certainty of a small net health benefit, with high certainty of at least
incremental net health benefit
C+=”Comparable or Better” - Moderate certainty of a comparable net health benefit, with high certainty of at
least comparable net health benefit
P/I = “Promising but Inconclusive” - Moderate certainty of a small or substantial net health benefit, small (but
nonzero) likelihood of a negative net health benefit
I = “Insufficient” – Either moderate certainty that the best point estimate of comparative net health benefit is
comparable or inferior; or any situation in which the level of certainty in the evidence is low
©Institute for Clinical and Economic Review, 2015
Page 15
Our search identified the same 25 trials that form the basis of two published systematic reviews and
meta-analyses of the safety and efficacy of PCSK9 inhibitors.1,2 We verified the data for their metaanalyses and primarily used the results of their meta-analyses to summarize the results from the
trials. When not available elsewhere, we used data from the publicly-available FDA briefing
documents and manufacturer submissions to the FDA for additional adverse event reporting.
For the meta-analyses, we followed the methodology reported in the meta-analysis by Navarese
and colleagues.1 Heterogeneity was assessed using the Cochran q test and the I2 statistic. If the
inconsistency was high (I2 ≥ 50%), then the results were combined using a random effects model.
Otherwise a fixed effects model was used. We assessed for publication bias using funnel plots and
Egger’s statistic.
We focused our analyses on the effect of PCSK9 inhibitors as a class because the LDL-C lowering
effects are similar, there are no head to head trials comparing them, and because of currently
limited data on their effects on key clinical outcomes, such as stroke, MI, and cardiovascular death.
Where possible, we looked at important subgroups of patients who might benefit from PCSK9
inhibitors. These include patients with HoFH, those with HeFH, those who have already experienced
CVD events, those without CVD events who are at high risk for events, and patients eligible for
statin therapy who are intolerant of statins. Unfortunately, most of the trials included a mix of
patients with and without prior CVD and did not report the results by those subgroups. The large
outcomes trials currently in progress (ODYSSEY OUTCOMES, FOURIER) are specifically enrolling only
patients with recent CVD events as this population is at highest risk for future events and thus will
be most likely to benefit from PCSK9 inhibitor therapy.
©Institute for Clinical and Economic Review, 2015
Page 16
4.2 Results
Study selection and Patient Population
The search identified 41 references describing 8 phase 2 trials, 16 phase 3 trials and one long-term
follow-up study.8,23-62 There are 13 trials (3 phase 2 trials) of alirocumab, including a total of 5,137
patients. Five of the trials of alirocumab have only been presented at conferences.23,31,34,44 There are
10 trials (5 phase 2 trials) of evolocumab including a total of 5,022 patients. In addition, the OSLER
trial re-randomized 4,465 participants from the phase 2 and 3 trials of evolocumab and followed
them for 52 weeks.56 The OSLER results are not included in the primary meta-analyses because the
participants are already included in the results of the primary trials. However, the OSLER results
(combined OSLER 1 and OSLER 2 trials) will be summarized because of the relatively long duration
of follow-up and large size of the trial. All of the trials of evolocumab have been published in the
peer-reviewed literature.
Appendix 6 Table 1 provides an overview of the studies. The TESLA Part B trial (evolocumab) is the
only trial that specifically enrolled patients with HoFH.47 The remaining trials enrolled patients with
either HeFH or non-specific hypercholesterolemia. The Gauss and Gauss 2 trials61,62 (evolocumab)
and the Odyssey Alternative trial 43 only enrolled patients with statin intolerance. There were no
trials that only enrolled patients with known CVD, although most of the trials had a significant
portion of participants with CVD. Follow-up was longer than one year for 5 trials, one year for 2
trials and less than one year for the remaining 17 trials. PCSK9 inhibitor therapy was compared to
placebo in 14 trials, to ezetimibe in 7 trials, and to both placebo and ezetimibe in the remaining 3
trials.
The average age of the participants was 51 to 66 years with the exception of the TESLA Part B trial
(HoFH patients) in which the participants had an average age of 31. About half of the participants
were female, approximately 30% had prior CVD, and approximately 15% had diabetes mellitus. High
intensity statin therapy was used in 17 of the 24 trials.
Quality of individual studies
The assessment of the quality of individual studies is summarized in Appendix 6 Table 2. For the
unpublished studies, peer reviewed publications27,35,43 of the study design and rationale were used
to supplement the abstracts and slide presentations. There was low risk of bias in all of the trials
except the OSLER 1 and 2 trials. Randomization was done appropriately with allocation
concealment and blinding of the participants, the investigators, and the staff performing outcome
adjudication. Follow-up retention was high and analyses were performed adhering to intention-totreat principles. All of the studies were funded by the manufacturers. In the OSLER trials,
appropriate randomization and allocation concealment was performed, but there was no blinding
of patients, investigators, or staff, so there was increased risk of bias.
©Institute for Clinical and Economic Review, 2015
Page 17
Key Studies
Alirocumab
The ODYSSEY LONG TERM trial is the largest randomized trial of alirocumab.50 The eligible
population included adults with known coronary heart disease (69%) or a coronary heart disease
(CHD) risk equivalent (peripheral artery disease, ischemic stroke, chronic kidney disease, or diabetes
with at least two additional risk factors). Patients also had an LDL-C level ≥ 70 mg/dL on statin
therapy. All participants continued to take high intensity statin therapy (47%) or the highest
tolerated dose (53%). The investigators randomized 2,341 participants in a 2:1 ratio to alirocumab
150 mg every two weeks (n=1530) or to identical placebo (n=780) and followed them for 78 weeks.
At 24 weeks LDL-C levels declined from 122.8 to 48.3 mg/dL in the alirocumab group and from
122.0 to 118.9 in the placebo group. The reduction in LDL-C was greater in the alirocumab group at
24 weeks (-61.0% vs. 0.8%, p<0.001) and at 78 weeks (-52.4% vs. 3.6%, p<0.001).
Overall, there was no trend towards more AEs, serious AEs, diabetes, liver enzyme elevation, or
muscle enzyme elevation in the alirocumab group (see Table 2 on the following page). There was a
trend towards more discontinuations due to AEs, more neurocognitive AEs, and more injection site
reactions, but these were not statistically significant, particularly if adjustment is made for multiple
comparisons. For the major cardiovascular adverse events, there were trends towards fewer deaths
from CHD (0.3% vs. 0.9%, p=0.26), fewer non-fatal MIs (0.9% vs. 2.3%, p=0.01), and less unstable
angina requiring hospitalization (0% vs. 0.1%, p=0.34), but more fatal and non-fatal ischemic strokes
(0.6% vs. 0.3%, p=0.35). There was a significant reduction in the sum of these four major adverse
cardiovascular events (MACE 1.7% vs. 3.3%, p=0.02). In the alirocumab group, the 575 participants
(37%) with LDL-C levels < 25 mg/dL on two or more consecutive measurements had similar rates of
adverse events as the overall alirocumab group.
The trial unequivocally demonstrates that alirocumab lowers LDL-C levels compared with placebo.
The study is underpowered to evaluate uncommon AEs but selected AEs can be found in Table 2 on
the next page. Given the large number of AEs evaluated, even those achieving statistical
significance at a p-value of 0.05 may be due to chance. For example, at 52 weeks follow-up in the
other large RCT of alirocumab26 (ODYSSEY COMBO II trial, n=720), there were more non-fatal MIs in
the alirocumab group (2.5% vs. 1.2%) and fewer patients reporting myalgias (4.4% vs. 5.0%), both in
the opposite direction of the ODYSSEY LONG TERM results. Larger clinical trials with longer followup are needed to adequately address the balance of benefits and harms for the PCSK9 inhibitors.
©Institute for Clinical and Economic Review, 2015
Page 18
Table 2: Selected adverse events in the ODYSSEY LONG TERM trial
Adverse event (AE)
Any AE
Serious AE
AE leading to drug
discontinuation
Death from CHD
Non-fatal myocardial
infarction (MI)
Stroke
Myalgias
Neurocognitive AE
New diabetes
Alanine Aminotransferase
(ALT) elevation
Creatine kinase (CK)
elevation
Injection site reaction
Alirocumab (%)
81.0
18.7
7.2
Placebo (%)
82.5
19.5
5.8
P value
0.40
0.66
0.26
0.3
0.9
0.9
2.3
0.26
0.01
0.6
5.4
1.2
1.8
1.8
0.3
2.9
0.5
2.0
2.1
0.35
0.006
0.17
0.84
0.75
3.7
4.9
0.18
5.9
4.2
0.10
Evolocumab
TESLA Part B
The TESLA Part B trial is unique among the trials because it randomized 49 participants ≥ 12 years
old who were diagnosed with HoFH.47 All participants were taking statins (94% high intensity) and
most were also taking ezetimibe (92%). Despite maximal lipid lowering therapy, the baseline LDL-C
level was 9.0 mmol/L (348 mg/dL). The participants were randomized to evolocumab 420 mg or
placebo every 4 weeks for 12 weeks. LDL-C decreased by 23.1% in the evolocumab group and
increased by 7.9% in the placebo group (between group difference: 30.9%, p<0.0001). The
percentage change in LDL-C is relatively low in this study, but 92% of patients were taking
ezetimibe. There were no deaths, no serious adverse events, and no discontinuations due to AEs.
Overall AEs were less common in the evolocumab group (36% vs. 63%). The trial was too small and
follow-up too short to evaluate clinical outcomes.
DESCARTES
The DESCARTES trial is the only primary randomized trial of evolocumab with greater than 12 weeks
of follow-up. The eligible population included adults with an LDL-C level ≥ 75 mg/dL. Approximately
15% of the participants had prior CVD and 12% had diabetes. The investigators randomized 901
participants in a 2:1 ratio to evolocumab 420 mg every four weeks (n=599) or to identical placebo
(n=302) and followed for 52 weeks. At 52 weeks LDL-C levels declined 50.6% from a baseline of
©Institute for Clinical and Economic Review, 2015
Page 19
100.4 mg/dL in the evolocumab group and increased 8.7% from 100.2 mg/dL in the placebo group
(between group difference: 57.0%, p<0.001).
Overall, there were no more AEs in the evolocumab group, as seen in Table 3 below. There were
nominally more serious AEs, discontinuations due to AEs, atherosclerotic events, myalgias, injection
site reactions and CK elevations, though the absolute differences were small. Neurocognitive AEs
were not reported.
Table 3: Selected adverse events in the DESCARTES trial
Adverse event (AE)
Any AE
Serious AE
AE leading to drug
discontinuation
Death
Atherosclerotic event
Myalgias
ALT elevation
CK elevation
Injection site reaction
Evolocumab (%)
74.8
5.5
2.2
Placebo (%)
74.2
4.3
1.0
P value
NR
NR
NR
0
1.0
4.0
0.8
1.2
5.7
0.3
0.7
3.0
1.0
0.3
5.0
NR
NR
NR
NR
NR
NR
NR Not reported
OSLER 1 and 2
The OSLER trials were open label randomized trials of the participants in the all of the Phase 2 trials
of evolocumab (OSLER 1) and all of the Phase 3 trials (OSLER 2) who agreed to be re-randomized for
extended follow-up.56 The results of the two trials were combined in the published results as the
OSLER trial. The investigators randomized 4,465 participants (74.1% of eligible participants) in a 2:1
ratio to evolocumab either 140 mg every two weeks or 420 mg once a month (n=2,976) or to
standard therapy without placebo (n=1489) and followed for 52 weeks. The baseline LDL-C level
was 120 in the evolocumab group and 121 in the standard therapy group. At 48 weeks the between
group difference in LDL-C was 58.4% (p<0.001).
Transient evolocumab-binding antibodies were detected in 0.3% of patients in both groups, but
were not neutralizing. As in the prior trials, the AEs are similar in the two groups (Table 4 on the
following page), though may be more subject to reporting bias because of the lack of blinding in the
OSLER trials.
©Institute for Clinical and Economic Review, 2015
Page 20
Table 4: Selected adverse events in the OSLER 1 AND 2 trials
Adverse event (AE)
Any AE
Serious AE
AE leading to drug
discontinuation
Death from CHD
Non-fatal MI
Stroke
Myalgia
Neurocognitive AE
New diabetes
ALT elevation
CK elevation
Injection site reaction
NR Not reported
Alirocumab (%)
69.2
7.5
2.4
Placebo (%)
64.8
7.5
NA
P value
NR
NR
NR
0.1
0.3
0.1
3.0
0.9
1.1
1.0
0.6
4.3
0.2
0.3
0.1
2.9
0.3
0.7
1.2
1.1
NA
NR
NR
NR
NR
NR
NR
NR
NR
NR
NA Not applicable
Clinical Benefits
The section that follows evaluates the effectiveness of PCSK9 inhibitors as a class, including the
percentage LDL-C lowering effects of PCSK9 inhibitors versus placebo and versus ezetimibe, and
differences by individual drug dose. Background lipid therapy may be important because statin
therapy raises PCSK9 levels in the blood, so the percentage reduction of LDL-C tends to be greater
in individuals taking statin therapy compared to those not receiving such therapy. Clinical
outcomes, including total mortality, CVD mortality, non-fatal MI, and stroke are then discussed. We
have elected to not apply any limits to the PCSK9 inhibitor evidence base used for this analysis. This
decision was made because no trials were designed with clinical events as the primary outcome, the
number of events is low, and the lipid lowering effects are similar for the approved doses of PCSK9
inhibitors. The meta-analysis results of Navarese and colleagues are used unless otherwise noted.1
LDL-C reduction
The summary estimate for percentage reduction in LDL-C is 47.5% (95% CI: 25.4 to 69.6). There was
no evidence for publication bias for the meta-analysis of LDL-C reduction (Egger’s p-value 0.99).
However, there was significant heterogeneity observed (I2 = 93%, p<0.001). As noted in the
Methods, an I2 > 50% is considered high. Sources of heterogeneity likely include differences in the
patient populations studied (baseline lipid levels, background statin therapy), differences in the
dose and type of PCSK9 inhibitor, and differences in the comparison group. These are explored
further in stratified analyses in the table below.
©Institute for Clinical and Economic Review, 2015
Page 21
Table 5: Meta-analysis of the percentage reduction in LDL-C by PCSK9 inhibitors as a class in
10,159 participants in phase 2 and 3 randomized trials stratified by background statin therapy
Comparison group
Placebo
Ezetimibe
All, % (95% CI)
58.8 (56.5 to 61.0)
36.2 (33.1 to 39.3)
Background Statin Therapy
No statin, %
High intensity
statin, %
53.6
57.9
36.2
34.4
Other statin, %
65.2
37.5
The percentage reduction in LDL-C is greater when PCSK9 inhibitors are compared to placebo
(58.8%) than that observed compared to ezetimibe (36.2%). The percentage reduction in LDL-C
varies much less by background statin therapy.
Findings are stratified by dose and type of PCSK9 inhibitor in Table 6 below. Evolocumab may have
slightly greater LDL-C reductions than alirocumab, but the differences are small compared to the
percentage reduction achieved by either of the PCSK9 inhibitors. Furthermore, differences in the
underlying populations studied may explain the relatively small differences in the percentage
reduction in LDL-C. The lack of head to head randomized trials makes it impossible to conclude that
one of the PCSK9 inhibitors lowers cholesterol more than the other.
Table 6: Meta-analysis of the percentage reduction in LDL-C by PCSK9 inhibitors in 10,159
participants in phase 2 and 3 randomized trials by stratified by dose and type of PCSK9 inhibitor
Comparison
group
Placebo
Ezetimibe
All, % (95% CI)
58.8 (56.5 to 61.0)
36.2 (33.1 to 39.3)
Dose and type of PCSK9 inhibitor
Alirocumab 75 Alirocumab 150
mg Q2W
mg Q2W
52.6
56.2
31.7
*
Evolocumab
140 mg Q2W
63.5
39.3
Evolocumab
420 mg Q4W
57.3
37.5
* Insufficient data
The PCSK9 inhibitors also improved other lipid parameters. HDL cholesterol increased by 6.1%
compared with placebo and 6.8% compared with ezetimibe (P<0.001 for both). There were also
significant reductions in total cholesterol (39% vs. placebo, 24% vs. ezetimibe, p<0.001 for both)
and in lipoprotein (a) (28% vs. placebo, 24% vs. ezetimibe, p<0.001 for both).
The evidence clearly shows that the PCSK9 inhibitors improve all lipid parameters significantly
compared to placebo or ezetimibe therapy. These findings are consistent whether the population
studied is taking high intensity statin therapy, lower intensity statin therapy, or no statin therapy at
baseline.
LDL-C lowering by patient subpopulation
HoFH: As described earlier, the TESLA Part B trial47 is the only trial that randomized patients with
known HoFH. The percentage LDL-C reduction of evolocumab compared to placebo was lower in
©Institute for Clinical and Economic Review, 2015
Page 22
this trial (30.9%) than seen in other trials. This may reflect the much higher pre-treatment LDL in
this population (356 mg/dL), or it may reflect lower efficacy of PCSK9 inhibitors in patients with one
or more of the mutations represented in the trial. In addition, 92% of the participants were taking
ezetimibe: the observed difference is similar to that observed in the studies that randomized
patients to either a PCSK9 inhibitor or ezetimibe. Even though the relative reduction in LDL was
lower than in the other trials, the absolute reduction (74 mg/dL) is similar to that reported in
populations with HeFH or non-specific hypercholesterolemia.
HeFH: 10 studies randomized primarily participants with HeFH. The LDL-C reduction in this
subpopulation ranged from 39.1% to 63.9% compared to placebo and from 34% to 35.8% compared
to ezetimibe. These subgroup findings are quite similar to the overall results.
Non-specific hypercholesterolemia (HC): 13 studies randomized primarily participants with nonspecific hypercholesterolemia. The LDL-C reduction in this subpopulation ranged from 45.9% to
70.9% compared to placebo and from 27.2% to 43.4% compared to ezetimibe. Again, these
subgroup findings are quite similar to the overall results.
Statin intolerant: Three studies randomized statin intolerant patients (GAUSS, GAUSS 2, ODYSSEY
ALTERNATIVE).44,61,62 In the GAUSS trial, there was a 47.3% LDL-C reduction with evolocumab
treatment compared to placebo, and a 35.9% reduction compared to ezetimibe. In the GAUSS 2
trial, the LDL-C reduction was 38.1% with every 2 week dosing and 37.6% with every 4 week dosing
both compared to ezetimibe. Finally, in the ODYSSEY ALTERNATIVE trial, the LDL-C reduction with
alirocumab treatment was 30.4% compared to ezetimibe.
The percentage LDL-C reduction in the statin intolerant patient population was similar to that of
both the FH and HC populations. There do not appear to be large differences between the LDL-C
lowering effects of PCSK9 inhibitors in HeFH, HC, or statin intolerant patient populations. The HoFH
population, which is a small population with extraordinarily high LDL-C levels, appears to have a
lower percentage reduction in LDL-C than the other populations. The study results were not
presented in subgroups defined by a prior history of CVD events, so no pooled estimates could be
made.
Clinical outcomes
There are large long-term (5-year or more) outcomes studies ongoing for both alirocumab
(ODYSSEY OUTCOMES, n=18,000, >5 years) and evolocumab (FOURIER, n=22,500, 5 years) that
should present initial results in 2017. The clinical outcomes below from the meta-analysis by
Navarese represent CVD events in trials designed with LDL-C lowering as the primary outcome and
are reported as adverse events. The most important clinical outcomes for lipid lowering therapy
include death from cardiovascular disease, MI, stroke, and unstable angina requiring
hospitalization. There was little statistical heterogeneity for each of the outcomes (I2 = 0%), so fixed
©Institute for Clinical and Economic Review, 2015
Page 23
effects models were used. Navarese and colleagues did not report the stroke outcomes, so we
performed our own meta-analysis for this outcome using the same analytic methods (Table 7,
below). Publication bias was not apparent for any outcome, either by examining funnel plots or
Egger’s p statistic.
Table 7: Meta-analysis results for patient-oriented outcomes
Outcome
All-cause
mortality
CVD Mortality
MI
Stroke
Unstable angina
OR (95% CI)
p
I2
N
Events PCSK9 group (%)
0.45 (0.23-0.86)
0.015
0%
10,159
19 (0.3%)
Events control
group (%)
21 (0.5%)
0.50 (0.23-1.10)
0.49 (0.26-0.93)
1.97 (0.69-5.65)
0.61 (0.06-6.14)
0.084
0.030
0.206
0.676
0%
0%
0%
0%
10,159
5,195
4,683
3,894
12 (0.2%)
19 (0.6%)
14 (0.5%)
1 (0.05%)
13 (0.3%)
19 (1.0%)
3 (0.2%)
1 (0.08%)
The findings of the meta-analysis suggest that the PCSK9 inhibitors reduce the odds of all-cause and
cardiovascular mortality by about 50%; however the total number of events is low and the
confidence intervals are wide. The odds ratio for stroke in the meta-analysis was twice as high in
the PCSK9 group, but the confidence interval is very wide and not statistically significant. There
were no significant differences in these results when stratified by comparison group (placebo,
ezetimibe), by PCSK9 inhibitor (alirocumab, evolocumab), or when adjusted for length of follow-up.
In sensitivity analyses excluding the data from studies not yet published in the peer-reviewed
literature, the conclusions are the same.
Harms
As seen in the earlier section that described the largest trials of PCSK9 inhibitors, there were no
large differences in the overall adverse event rates between these agents and their comparators.
Meta-analyses of selected adverse event rates are summarized in Table 8 on the following page. We
focused on serious adverse events including those leading to drug discontinuation, adverse events
associated with other lipid lowering drugs (myalgias, neurocognitive events, liver and muscle
enzyme elevations), and those associated with monoclonal antibody injections (injection site
reactions, hypersensitivity reactions).
©Institute for Clinical and Economic Review, 2015
Page 24
Table 8: Meta-analysis results for selected harms
OR (95% CI)
P
I2
N
Serious AE
AE leading to drug
discontinuation
Myalgias
Neurocognitive AE
ALT elevation
1.01 (0.87-1.18)
1.03 (0.84-1.26)
0.879
0.773
0%
0%
10,159
9424
1.16 (0.91-1.49)
1.08 (0.57-2.06)
0.82 (0.54-1.24)
0.236
0.816
0.350
37%
13
0%
6269
6601
9108
CK elevation
Injection site reaction
Hypersensitivity reactions
0.72 (0.54-0.96)
1.30 (1.03-1.65)
0.69 (0.23-2.08)
0.026
0.029
0.510
0%
4.5%
3.5%
10,159
9028
1062
Outcome
Events PCSK9
group (%)
573 (9.3)
270 (4.7%)
Events control
group (%)
307 (7.7%)
167 (4.5%)
199 (5.1%)
27 (0.6%)
55 (1.0%)
108 (4.5%)
14 (0.6%)
39 (1.1%)
121 (2.0%)
222 (4.1%)
7 (1.2%)
92 (2.3%)
106 (3.0%)
7 (1.4%)
These meta-analysis results do not suggest that the PCSK9 inhibitors lead to elevations in serious
adverse event rates. There are more injection site reactions, which may lead to slightly higher rates
of drug discontinuation compared to the control group. There is a slight excess of neurocognitive
events with PCSK9 inhibitors, but the results are not statistically significant. There is also a trend
towards more myalgias in the PCSK9 treated participants, but this is balanced by a statistically
significant reduction in the number of participants with elevations in the muscle enzyme CK.
The results were generally consistent, and there was little evidence for publication bias except for
the neurocognitive adverse events, which were not consistently reported in the published studies.
Excluding data from the unpublished studies did not significantly impact the results.
Although these results do not identify any worrisome or unexpected adverse events, many of the
trials were 6 months or less in duration. Serious adverse events may be identified in the large 5-year
outcome trials that are currently in progress (ODYSSEY OUTCOMES, FOURIER).
4.3 Summary and Comment
Our analyses demonstrate that the existing evidence provides moderate certainty that PCSK9
treatment provides an incremental or substantial net health benefit for all of the patient
subpopulations included in the scope of this review. There is no question that the drugs improve
intermediate risk factors for cardiovascular disease. They substantially reduce LDL-C, total
cholesterol, and lipoprotein (a), and also modestly elevate HDL-cholesterol. A high-quality metaanalysis found a 50% reduction in all-cause mortality that was statistically significant and reductions
of similar magnitude (albeit not statistically significant) in death from cardiovascular disease and in
MIs.
The drugs also appear to be very well-tolerated. The randomized trials do not demonstrate an
increase in AEs, serious AEs, or drug discontinuations due to AEs. Neurocognitive event rates are
©Institute for Clinical and Economic Review, 2015
Page 25
low and do not appear to be increased in patients randomized to PCSK9 inhibitors compared to the
control patients.
However, there are several limitations in the evidence base that give reason for caution. There are
theoretical concerns that long term exposure to very low levels of cholesterol may have unexpected
adverse effects that have not been observed in the evidence base to date because the majority of
the studies lasted less than 6 months. In addition, as noted earlier, medications such as torcetrapib
that lower LDL-C, raise HDL, and have strong biological plausibility, have demonstrated increased
cardiovascular event rates and total mortality in long term studies. The large randomized trials with
long-term follow-up that are designed to evaluate the effect of the PCSK9 inhibitors on hard clinical
endpoints have completed recruitment, but their results will not be available until 2017.
The promising evidence on patient-centered outcomes from the published meta-analysis is also
limited in several ways. First, the 95% confidence intervals for the odds ratios estimating clinical
benefit either include 1.0 or approach 1.0. Second, the evidence in this meta-analysis combines
data from trials of two different PCSK9 inhibitors, each with two different dosing schedules, with
too few events in the evidence base to attempt subgroup analyses. Another limitation of the metaanalysis is that the populations studied were quite different: young adults with homozygous FH and
very high LDL-C; older adults with LDL-C < 100, but not at goal; and older adults who have already
had a heart attack or stroke. A final reason for caution about the findings for the meta-analysis is
that the PCSK9 inhibitors were compared to two different control arms: placebo and ezetimibe. The
percentage LDL-C reduction consistently favored PCSK9 inhibitors, but the magnitude varied slightly
by population and significantly by control group. It is likely that the clinical benefits will vary by
dose, drug, background drug therapy, and population studied.
Despite these limitations, the evidence base provides high certainty that PCSK9 inhibitors lead to
superior reductions in LDL-C levels compared to both placebo and ezetimibe. The percent reduction
in LDL-C with PCKS9 inhibitor treatment is approximately 55-60% and appears not to differ
substantially across different patient subpopulations. The potential net health benefit from this
level of LDL-C reduction will be greater among patient subpopulations at higher risks of CVD. Among
the subgroups, the population with HoFH is at highest risk for CVD events. Untreated, they have
CVD events in the second decade of life. Differences in CVD risk are less marked between patients
with HeFH and those with a prior history of CVD who have elevated LDL-C levels despite other
treatment and/or who cannot take statins.
In summary, the ICER review team believes that the existing evidence suggests, with moderate
certainty, that the net health benefit of the PCSK9 inhibitors is either incremental or substantial for
the patients in the subpopulations within the scope of this review. Despite the uncertainty in the
actual level of net health benefit, we believe there is less than a 10% chance that ongoing trials will
demonstrate a net harm from PCSK9 inhibitor treatment, and, therefore, our evidence rating within
the ICER Integrated Evidence Rating framework is “Promising but Inconclusive.”
©Institute for Clinical and Economic Review, 2015
Page 26
5. Other Benefits or Disadvantages
Our reviews seek to provide information on other benefits or disadvantages offered by the
intervention to the individual patient, caregivers, the delivery system, other patients, or the public
that would not have been considered as part of the evidence on comparative clinical effectiveness.
Examples include, but are not limited to:

Methods of administration that improve or diminish patient acceptability and adherence

A public health benefit, e.g. reducing new infections

Treatment outcomes that reduce disparities across various patient groups

More rapid return to work or other positive effects on productivity (if not considered a
benefit as part of comparative clinical effectiveness)

New mechanisms of action for treatments of clinical conditions (e.g., mental illness) for
which the response to currently available treatments varies significantly among patients for
unknown reasons (substantial heterogeneity of treatment effect)
Currently available PCSK9 inhibitors must be injected. This is a potential disadvantage compared to
most pharmaceuticals because some patients are unable to self-inject or experience anxiety
associated with self-injection. On the other hand, patients rapidly learn to inject themselves with
low molecular weight heparin and with insulin when needed, so the barrier may not be too high,
particularly for patients motivated by FH or a history of CVD events. Furthermore, the need to inject
the medication only once or twice a month may enhance adherence and be considered an
advantage compared to medications that need to be taken on a daily basis.
There do not appear to be other benefits or disadvantages of note to PCSK9 inhibitor therapy.
©Institute for Clinical and Economic Review, 2015
Page 27
6. Comparative Value
6.1 Overview
In order to assess the incremental costs per outcomes achieved of PCSK9 inhibitors, we conducted a
cost-effectiveness analysis using a previously validated model of cardiovascular disease in the
contemporary adult population of the United States (see section 6.2).63-65 We modeled the addition
of ezetimibe and PCSK9 inhibitors to background statin therapy as currently being used in the
population and examined the impact on MI, stroke, and cardiovascular death. We estimated drug
costs based on current prices, predicted population-level reductions in clinical outcomes, and
cardiovascular disease costs (hospitalizations, procedures, and chronic disease care costs) due to
the LDL-C lowering effects of the drugs to estimate the incremental cost-effectiveness of PCSK9
inhibitors and their anticipated budgetary impact. Clinical trials suggest only minor differences in
effectiveness between alirocumab and evolocumab based on frequency of dosing and impact on
LDL-C; therefore, we model PCSK9 inhibitors as a class for the purpose of this analysis.
Outputs from this model were also used to inform a population-based analysis of the one- and fiveyear budgetary impact of PCSK9 inhibitors, by key subpopulation and on an overall basis (see
section 6.3). Budgetary impact was assessed using assumed levels of uptake over these timeframes,
and included assessment of drug costs as well as cost savings from averted cardiovascular events.
We also define a “value-based price benchmark” for PCSK9 inhibitors based on a calculated
threshold for policy intervention to manage the costs of new pharmaceuticals.
6.2 Incremental Costs per Outcomes Achieved
Cost-Effectiveness Model: Methods
Model Structure
The CVD Policy Model is a computer-simulation, discrete-state Markov model of coronary heart
disease and stroke incidence, prevalence, mortality, and costs in the U.S. population over age 35.6365
The model was created at Harvard University in 1984 and has been used for more than 30 years
to provide evidence on the value of cardiovascular disease prevention approaches in U.S. adults.
The CVD Policy Model team has published reports from a number of high-impact studies of public
health and clinical interventions.66-75 The last model software and input data update was completed
in 2015.
The Demographic-Epidemiologic Sub model predicts coronary heart disease and stroke incidence
and non-CVD mortality among subjects without CVD, stratified by age, sex, and up to 8 additional
categorized risk factors estimated from weighted United States National Health and Nutrition
©Institute for Clinical and Economic Review, 2015
Page 28
Examination Surveys data from 2007-2010 (Figure 3). Risk factors include: systolic blood pressure
(<130, 130-139.9, ≥140 mm Hg), smoking status (active smoker, non-smoker with exposure to
environmental tobacco smoke, non-smoker without environmental exposure), high density
lipoprotein (HDL) cholesterol (<1.0, 1.0-1.5, ≥1.6 mmol/L; <40, 40-59.9, ≥60 mg/dL), low-density
lipoprotein cholesterol (LDL-C) (<2.6, 2.6-3.3, ≥3.4 mmol/L; <70, 70-99.9, ≥100 mg/dL), body mass
index (<25, 25-29.9, ≥30 kg/M2), diabetes mellitus (Type 1 or Type 2; yes or no), statin use (yes or
no). After CVD develops, the Bridge Sub model characterizes the initial stroke or coronary heart
disease event (cardiac arrest, MI, or angina) and its sequelae (including CVD mortality) for 30 days.
Then, the Disease History Sub model predicts subsequent CVD events, coronary revascularization
procedures, CVD mortality, and non-CVD mortality among patients with CVD, stratified by age, sex,
and history of events. The general chronic CVD categories are coronary heart disease only, stroke
only, and combined prior coronary heart disease and prior stroke. Each state and event has an
annual cost and quality-of-life adjustment, as well as an annual probability of a repeat event and/or
transition to a different CVD state. All population distributions, risk factor levels, coefficients, event
rates, case fatality rates, costs, and quality-of-life adjustments can be modified for forecasting
simulations.
Figure 3: Cardiovascular Disease Policy Model structure and disease states.
©Institute for Clinical and Economic Review, 2015
Page 29
We modeled the entire population of U.S. adults aged 35 to 74 years in the year 2015. We assumed
the health system perspective,76 considering all direct and induced medical costs and relevant
clinical outcomes over a lifetime analytic horizon (defined as until patients reach 95 years of age
due to the absence of high-quality epidemiologic data in older populations). Utilities and costs were
assigned to each clinical event in annual cycles and discounted at 3% annually.107 We conducted
extensive deterministic and scenario-based sensitivity analyses to account for uncertainty in the
input parameters. We adhered to the recommendations of the Panel on Cost -Effectiveness in
Health and Medicine where practicable.108 Additional modeling details, including sources of input
parameters and model calibration, are presented in Appendix A7.
With regards to the impact of statin therapy on LDL-C, we assumed a “flat” beta across age-groups
as demonstrated in clinical trials: i.e., that risk reduction in cardiovascular events per unit reduction
in cholesterol is identical in all age groups.109 The effect of LDL-C lowering on CHD prevention
assumed by the CVD Policy Model (relative risk per mg/dL LDL-C reduction) was validated in a
simulation of West of Scotland Coronary Prevention Study.71,110 Simulations of the US population
aged 45-64, imposing the pre- and post-intervention LDL-C and HDL cholesterol levels recorded in
the West of Scotland Study 9110 produced estimates of key clinical outcomes, i.e., cumulative CHD
mortality or first MI, and ratio of events in participants treated with statins or placebo, within 1% of
the numbers observed in the trial (Appendix 7 Table 2).
Importantly, while categorical definitions of LDL-C are used for the purpose of stratifying patients
without a CVD history into risk factor groups, continuous measures of LDL-C were used to assess
levels of risk reduction for the treatments of interest in our model.
The CVD Policy Model is written in Lahey Fortran 95. Processing of modeled outcomes was carried
out using QuickBasic64 and Excel 2011 (Microsoft, Redmond, Washington); statistical analyses were
performed using SAS 9.4 (SAS, Inc., Cary, North Carolina) and Stata 13 (StataCorp, College Station,
Texas).
©Institute for Clinical and Economic Review, 2015
Page 30
Target Population
In the base case, we evaluated the cost-effectiveness of PCSK9 inhibitors in three target
populations. Populations were chosen to approximate those described in the FDA-labeled
indications for alirocumab, which was approved in July 2015.77 This is in line with the idea that
because statins are both inexpensive and effective, PCSK9 inhibitors will probably be first used
among patients at highest risk for adverse cardiovascular events (Table 9 on the following page).
Three target populations:
1. Familial Hypercholesterolemia (FH)
At least three different definitions of FH are used in clinical practice; all clinical definitions
relate to high baseline levels of LDL-C and personal or family history of premature
coronary heart disease. 111,112 For the purpose of this analysis, we defined FH as a baseline
LDL-C level greater than or equal to 250mg/dL (6.465 mmol/L) among patients not on
statin therapy, and an LDL-C level greater than or equal to 200mg/dL (5.172 mmol/L)
among patients receiving statin therapy. We were unable to specifically identify patients
with a family history of premature coronary disease because of limitations of available
epidemiological data. For the purpose of this analysis, we assumed that 10% of the
population would be statin-intolerant, defined as being unable to tolerate even low-dose
statins, and varied this proportion between 3% and 20% in sensitivity analyses.113-115
2. Patients with pre-existing CVD (defined as a prior history of angina, MI, or stroke) whose
LDL-C level is greater than or equal to 70mg/dL (1.810 mmol/L) and who are unable to
tolerate statin therapy.
For the purpose of this analysis, we assumed that 10% of the population with pre-existing
CVD is statin-intolerant, defined as being unable to tolerate even low-dose statins. We
varied the prevalence of statin intolerance in sensitivity analyses between 3% and 20%.
3. Patients with pre-existing CVD (defined as a prior history of angina, MI, or stroke) whose
LDL-C level is greater than or equal to 70mg/dL (1.810 mmol/L) despite receiving
maximally tolerated statin therapy.
Demographic and key clinical characteristics of the target populations are shown in Appendix 7
Table 3.
©Institute for Clinical and Economic Review, 2015
Page 31
Table 9. Treatment Strategies Evaluated in this Report.
Treatment Strategies
We modeled three treatment strategies in patients able to tolerate statins, as shown in Table 9,
above:

background treatment with a statin (as treated in the population, control),

incremental treatment with ezetimibe among patients already on a statin, or

incremental treatment with a PCSK9 inhibitor among patients already on a statin.
In the base case, 10% of the population was deemed statin-intolerant. Where relevant, the
treatment strategies available to these patients were:

no treatment with lipid lowering therapies (control),

treatment with ezetimibe, or

treatment with a PCSK9 inhibitor.
In other words, the FH and CVD populations included patients who were receiving statin therapy,
those who could tolerate statins but were not receiving them, and patients who were intolerant of
statins (“background therapy with statins as tolerated”). In the base-case analyses for each
population, only patients who were either already receiving statin therapy or deemed statin
intolerant (10% of the population) received incremental therapy with either ezetimibe or a PCSK9
inhibitor (Table 10).
We estimated the degree of LDL-C reduction with maximally tolerated doses of ezetimibe and
PCSK9 inhibitors (when used alone or in combination with statins) from the published literature
(Tables 10 and 11). We assumed that the drugs were equally efficacious in all patient populations,
i.e., the proportion of reduction in LDL-C from baseline was constant across all subgroups studied.
©Institute for Clinical and Economic Review, 2015
Page 32
Our review of the literature yielded the following estimates of the effect of ezetimibe and PCSK9
inhibitors on LDL-C (Table 10):
Table 10. Effect of ezetimibe and PCSK9 inhibitors on serum LDL-C when used alone or
incremental effect when added to statins.
Medication
Background statin use
95% CI for
sensitivity analysis
Reference
None
Mixed low- and high-intensity
High-intensity
%LDL-C reduction
(incremental to statin
effect in statintreated)
53.65
65.24
57.93
PCSK9
inhibitor
47.78 – 5 9.51
60.02 – 70.46
54.91 – 60.95
1
Ezetimibe
None
Mixed low- and high-intensity
High-intensity
18.56
23.60
23.60
17.44 – 19.68
21.70-25.60
21.70-25.60
116
Abbreviations: Abbreviations: CI, confidence interval; LDLC, low-density lipoprotein cholesterol; PCSK9,
proprotein convertase subtilisin/kexin type 9.
Our review suggested the following effect of treatment with statin use on cardiovascular outcomes
(Table 11):
Table 11. Effect of ezetimibe and PCSK9 inhibitors on Clinical Outcomes (MI, stroke, and CV
death) per 1mmol/L reduction in LDL-C.
Medication
Risk Ratio for MI
and CV death
Risk ratio for
stroke
per 1 mmol/L
reduction in LDL-C
per 1 mmol/L
reduction in LDL-C
Statins
0.76
0.85
Ezetimibe
PCSK9
inhibitors
0.76
0.76
0.85
0.85*
Comment
Reference
We assumed a “flat” beta across agegroups as demonstrated in clinical
trials: i.e., that risk reduction in
cardiovascular events per unit
reduction in cholesterol is identical in
all age groups.
Since the effect of PCSK9 inhibitors on
stroke risk is not known, we
performed a sensitivity analysis that
assumed no effect on stroke.
109
109
1,109
Abbreviations: CI, confidence interval; CV, cardiovascular; LDLC, low-density lipoprotein cholesterol; MI, myocardial
infarction; PCSK9, proprotein convertase subtilisin/kexin type 9.
* Assumed in the base-case analysis.
©Institute for Clinical and Economic Review, 2015
Page 33
We assumed that these drugs affect cardiovascular outcomes (non-fatal MI, non-fatal stroke, and
cardiovascular death) in proportion to their effect on LDL-C: for one unit decline in LDL-C, we
assumed that statins, ezetimibe, and PCSK9 inhibitors reduce the risk of non-fatal MI and
cardiovascular death by an identical amount. In the absence of long-term effectiveness data, it is
uncertain whether LDL-C reduction with PCSK9 inhibitors produces similar reductions in clinical
events compared with statins. On the one hand, results of short-term clinical trials suggest greater
reductions in CV mortality and MI rates compared with statin treatment trials or observational
cohort studies for the same unit difference in LDL-C, so that theoretically, PCSK9 inhibitors may
have better clinical outcomes than with statins. On the other hand, statins have been ascribed
“pleiotropic” effects that may stabilize an atherosclerotic plaque and lower risk beyond that
expected for the same unit difference in LDL-C in observational cohort studies—and it may be that
PCSK9 inhibitors do not provide these additional benefits. In light of this uncertainty, we performed
a one-way sensitivity analysis in which we assessed the impact of a 25% greater or lesser effect of
PCSK9 inhibitors on MACE compared with equivalent LDL-C reduction observed in statin treatment
trials.
In the base-case, we assumed that statins, ezetimibe, and PCSK9 inhibitors also reduce the risk of
stroke (mediated by their effect on ischemic stroke but incorporated in the model as the adjusted
effect on total stroke). Previous studies examining the impact of lipid-lowering agents on stroke
have yielded mixed results (see discussion below). Furthermore, as noted in the review of clinical
evidence presented above, the effect of PCSK-9 inhibitors on stroke is uncertain. We therefore
performed a sensitivity analysis that assumed that treatment with PCSK9 inhibitors does not lower
the risk of stroke.
In clinical trials, serious adverse events with treatment with ezetimibe or PCSK9 inhibitors were
infrequent and did not statistically differ across treatment arms.50,56,78 For this analysis, we did not
model costs or disutilities associated with these minor adverse events such as injection site
reactions.
Note, however, that long-term effectiveness or safety data are not presently available for PCSK9
inhibitors.
Costs
Age- and sex-specific health care costs were estimated using national data.71 Hospitalized stroke
and coronary heart disease costs and acute stroke rehabilitation costs were estimated using
California hospital data117 and deflated using cost-to-charge ratios118 and the ratio of the U.S.
national average costs to the California average.119 Chronic outpatient CVD costs additional to
average background health care costs for the first year after an event and for subsequent years
were estimated for patients with a stroke or coronary heart disease diagnosis surveyed in the U.S.
Medical Expenditure Panel Surveys (MEPS) pooled from 1998-2008. Average annual non©Institute for Clinical and Economic Review, 2015
Page 34
cardiovascular (background) costs were also estimated from the MEPS.120 All model costs were
indexed to the year 2015 using the medical component of the U.S. Consumer Price Index.94 We
assumed the annual cost of ezetimibe to be $2,828, based on the wholesale acquisition cost
(WAC);50 costs of atorvastatin 80 mg (used in one of the scenario analyses) were estimated to total
$812 per year, based on median WAC across brand and generic versions.78 We assumed the annual
cost of PCSK9 inhibitors to be equal to the average of the recently announced annual wholesale
prices of alirocumab ($14,600 per patient per year) and evolocumab ($14,100 per patient per
year).25 Drug costs were subjected to a variety of sensitivity and threshold analyses.
Utilities
Health-related quality-of-life weights and severity distributions for CVD disease states were based
on the Global Burden of Disease disability weights study.121-123
Outcome Measures
We report results in 2015 U.S. dollars, quality-adjusted life years (QALYs), and incremental costeffectiveness ratios (ICERs).108 We assessed the cost-effectiveness of both PCSK9 inhibitors and
ezetimibe relative to comparator treatment (i.e., statin therapy in those who could tolerate it and
no therapy in those who could not).
ICER =
Cost Therapy being evaluated – Cost Next most effective therapy
Effectiveness Therapy being evaluated − EffectivenessNext most effective therapy
In the base case, we only included costs related to drug therapy and all costs related to the
management of cardiovascular disease. In a sensitivity analysis, we also included the costs related
to management of other conditions. We assumed a willingness-to-pay threshold of $100,000/QALY.
We also report the number of patients that would need to be treated for five years (NNT5) to avert
one major adverse cardiovascular event (MACE, defined as cardiovascular death, nonfatal MI, or
nonfatal stroke).
Sensitivity Analyses
Sensitivity analyses examine the impact of uncertainty in various input parameters on the estimates
of cost-effectiveness of the therapies being examined in the model.
We performed various one-way sensitivity analyses for each of the target subpopulations (Table 12)
by varying one input parameter at a time while holding all other parameters constant at their basecase values. In particular, there is uncertainty in extrapolating the impact of statins on stroke to
PCSK9 inhibitors because of the following:

Observational studies demonstrated no association between LDL-C and stroke
©Institute for Clinical and Economic Review, 2015
Page 35

Randomized trials of statins show a decrease in stroke rates with statins compared with
placebo, for high intensity statins vs. low intensity statins, and for ezetimibe relative to
placebo 78,109,124

Randomized trials of other cholesterol-lowering agents have demonstrated either no impact
on stroke (fibrates, resins, diet) or an increase in stroke (hormone therapy/estrogens).
In light of this uncertainty, we performed a sensitivity analysis that assumed that treatment with
PCSK9 inhibitors has no effect on the risk of stroke (Table 12).
Table 12. Upper and lower bounds of inputs explored in one-way sensitivity analyses.
Target Subpopulation
Familial Hypercholesterolemia
History of CVD, statin
intolerant
History of CVD, on
statin therapy
One-Way Sensitivity Analyses
LDL-C lowering by PCSK9 (base case and range as reported in Table 10)
PCSK9 inhibitor effect on CV outcomes relative to statins per mmol/L
reduction in LDL-C (base case = 1 [equivalent to statins], range: 0.25-1.25)
Stroke benefit for PCSK9 (base case RR = 0.85 per mmol/L reduction in
LDL-C, range: 0.85-1.00)
Inclusion of non-cardiovascular costs into the ICER (base case = 0% of noncardiovascular costs included; range: 0-100%)
Prevalence of statin-intolerance (base case = 10%, range: 3-20%)
Analytic horizon (base case = lifetime, range: 20 years - lifetime)
Drug cost for PCSK9 inhibitors (base case = $14,350 per patient per year;
range: 50-200% of base-case)
LDL-C lowering by PCSK9 (base case and range as reported in Table 10)
PCSK9 inhibitor effect on CV outcomes relative to statins per mmol/L
reduction in LDL-C (base case = 1 [equivalent to statins], range: 0.75-1.25)
Stroke benefit for PCSK9 (base case RR = 0.85 per mmol/L reduction in
LDL-C, range: 0.85-1.00)
Inclusion of non-cardiovascular costs into the ICER (base case = 0% of noncardiovascular costs included; range: 0-100%)
Analytic horizon (base case = lifetime, range: 20 years - lifetime)
Drug cost for PCSK9 inhibitors (base case = $14,350 per patient per year;
range: 50-200% of base-case)
LDL-C lowering by PCSK9 (base case and range as reported in Table 10)
PCSK9 inhibitor effect on CV outcomes relative to statins per mmol/L
reduction in LDL-C (base case = 1 [equivalent to statins], range: 0.75-1.25)
Stroke benefit for PCSK9 (base case RR = 0.85 per mmol/L reduction in
LDL-C, range: 0.85-1.00)
Inclusion of non-cardiovascular costs into the ICER (base case = 0% of noncardiovascular costs included; range: 0-100%)
Analytic horizon (base case = lifetime, range: 20 years - lifetime)
Drug cost for PCSK9 inhibitors (base case = $14,350 per patient per year;
range: 50-200% of base-case)
Source
1
Assumed
1,109
Assumed
113-115
Assumed
25
, Range
Assumed
1
Assumed
1,109
Assumed
Assumed
25
, Range
Assumed
1
Assumed
1,109
Assumed
Assumed
25
, Range
Assumed
Abbreviations: CVD, cardiovascular disease; FH, familial hypercholesterolemia; ICER, incremental cost-effectiveness ratio; LDL-C,
low-density lipoprotein cholesterol; PCSK9, proprotein convertase subtilisin/kexin type 9.
©Institute for Clinical and Economic Review, 2015
Page 36
We also performed several scenario analyses, varying combinations of parameters in a clinically
meaningful manner (Table 13).
Table 13. Scenario Analyses.
Target
Subpopulation
FH
History of CVD,
high-risk
Scenario Analyses
In the base case, only patients who met the operational definition of FH and were
either already receiving statin therapy or were deemed statin intolerant (10% of the
population) received incremental therapy with ezetimibe or a PCSK9 inhibitor. In a
scenario analysis, we evaluated the impact of “full treatment” in which all statintolerant patients who were not already receiving statins were first treated with highintensity statins, after which the entire FH subpopulation was incrementally treated
with ezetimibe or a PCSK9 inhibitor.
In the base case, all patients with pre-existing CVD and LDL-C ≥70mg/dL on statin
therapy received incremental treatment with ezetimibe or a PCSK9 inhibitor. In a
scenario analysis, we evaluated the effect of only initiating therapy after an incident
MI. In this analysis, all patients who had an incident (first-ever) MI in 2015 received
incremental PCSK9 inhibitor therapy and were followed over their lifetime.
Abbreviations: CVD, cardiovascular disease; FH, familial hypercholesterolemia; MI, myocardial infarction; PCSK9, proprotein
convertase subtilisin/kexin type 9.
Finally, in pre-specified threshold analyses, we evaluated the price at which PCSK9 inhibitors would
be considered cost-effective at willingness-to-pay thresholds of $50,000/QALY, $100,000/QALY, and
$150,000/QALY.
Cost-Effectiveness Model: Results
Familial Hypercholesterolemia
The operational definition of FH in our model (baseline LDL-C level greater than or equal to
250mg/dL [6.465 mmol/L] among patients not on statin therapy and an LDL-C level greater than or
equal to 200mg/dL [5.172 mmol/L] among patients receiving statin therapy) identified 605,000
patients with FH, equating to 13.3 million patient-years of treatment over twenty years. The risk of
MACE in this population was 2.2-3.4 fold higher than an age- and gender-matched population that
does not meet the definition of FH and does not have pre-existing CVD. Key results are reported in
Table 14, one-way sensitivity analyses are shown in Figure 3, and additional results are shown in
Appendix 7 Table 5.
Compared with the control arm, incremental treatment with ezetimibe would avert 115,900 MACE
over the lifetime analytic horizon and produce 250,600 additional QALYs with an ICER of
$135,000/QALY compared with current treatment. Compared with the control arm, adding PCSK9
inhibitors to current treatment averted 324,200 MACE and produced 665,200 additional QALYs,
producing an ICER of $290,000/QALY. This higher ICER for PCSK9 inhibitors was driven by
©Institute for Clinical and Economic Review, 2015
Page 37
differences in drug costs ($14,350 per year for PCSK9 compared with $2,828 per year for ezetimibe).
We did not model HoFH separately, because the expected number of patients is small (n=300-400
in the US).
Table 14. Base-Case Clinical and Economic Outcomes Among Patients with FH.*
Personyears of
treatment
(millions)
comparator
Total
MACE
averted
NNT5†
QALYs
gained⌃
Incremental
Drug Costs⌃
(million $)
Incremental
Costs, Other
CV Care⌃
(million $)
ICER
($/QALY)
Statin +
Ezetimibe||,¶
22.3
115,900
77
250,600
$40,359
-$6,632
$135,000
Statin + PCSK9
inhibitor**,¶
23.7
324,200
28
665,200
$210,516
-$17,304
$290,000
Statin§
Abbreviations: CV, cardiovascular; FH, familial hypercholesterolemia; ICER, incremental cost-effectiveness ratio; MACE, major
adverse cardiovascular event (nonfatal MI, nonfatal stroke, and cardiovascular death); NNT5, number-needed-to-treat; QALY,
quality-adjusted life year.
* In the base case, all patients who met the operational definition of FH and were either already receiving statin therapy or
deemed statin-intolerant (10% of the population) received incremental therapy with ezetimibe or a PCSK9 inhibitor (n =
605,000 in 2015). The analytic horizon was lifetime (defined as when patients reach the age of 95 years). To reflect the
precision of the model, person-years of treatment are rounded to the nearest 100,000s; MACE and QALYs are rounded to the
100s; costs are rounded to the millions; and ICERs to the 1000s.
† Number of patients that would need to be treated for 5 years to avert one MACE event.
⌃ All costs are reported in 2015 U.S. dollars. Future costs and QALYs are discounted 3% a year.
§ Patients deemed to be statin-intolerant (base-case prevalence = 10% of the FH population) received no lipid-lowering
therapy.
|| Patients deemed to be statin-intolerant (base-case prevalence = 10% of the FH population) received only ezetimibe.
¶ Both statin+ezetimibe and statin+PSCK9 inhibitor arms are compared with the statin-only arm.
** Patients deemed to be statin-intolerant (base-case prevalence = 10% of the FH population) received only a PCSK9 inhibitor.
©Institute for Clinical and Economic Review, 2015
Page 38
One-Way Sensitivity Analyses:
In one-way sensitivity analyses, we examined the impact of varying individual input parameters
while holding all others constant (Figure 4 below). The ICER was very sensitive to changes in drug
cost and the analytic horizon, with higher drug costs and shorter analytic horizons reducing the
cost-effectiveness of the therapy.
Figure 4. One-way sensitivity analyses among patients with FH.
Base-case ICER = = $290,000/QALY. The ICER was very sensitive to changes in drug cost and the analytic horizon, with higher
drug costs and shorter analytic horizons lowering the cost-effectiveness of the therapy. Abbreviations: CVD, cardiovascular
disease; ICER, incremental-cost-effectiveness ratio; LDL-C, low-density lipoprotein cholesterol; PCSK9, proprotein convertase
subtilisin/kexin type 9; QALY, quality-adjusted life year.
* In this sensitivity analysis, we vary the effect of LDL-C reduction by a PCSK9 inhibitor relative to an equivalent reduction in
LDL-C by a statin. The base case assumes that a 1mmol/dL reduction in LDL-C by a PCSK9 inhibitor produces the same effect on
CVD outcomes as a 1mmol/dL reduction in LDL-C by a statin; this one-way analysis increases or decreases the relative effect of
PCSK9 inhibitors on CVD outcomes by 25%.
©Institute for Clinical and Economic Review, 2015
Page 39
Scenario Analysis: Initiating Therapy with Ezetimibe or PCSK9 After “Full Treatment” with Statins
in the FH population.
In a scenario analysis, we evaluated the impact of “full treatment” in which all patients who could
tolerate statins but were not already receiving statins at baseline were first treated with highintensity statins, after which the entire FH subpopulation was incrementally treated with ezetimibe
or a PCSK9 inhibitor. Key results are presented in Table 15 below; additional details are presented in
Appendix 7 Table 6. ICERs for the addition of ezetimibe or PCSK9 inhibitors are higher in this
analysis than in the base case due to the clinical benefit of moving untreated FH patients onto highintensity statin therapy; in fact, the “full treatment” strategy is cost-saving relative to the base case
as-treated approach.
Table 15. Scenario Analysis: Clinical and Economic Outcomes Assuming “Full Treatment” of FH
Patients with Statins Prior to Incremental Treatment with Ezetimibe or a PCSK9 inhibitor.*
Statin (as
treated)§
Statin (full
treatment)§,||
PersonTotal
years of
MACE
treatment
averted
(millions)
comparator
NNT5†
QALYs
gained⌃
Incremental
Drug Costs⌃
(million $)
Incremental
Costs, Other
CV Care⌃
(million $)
ICER
($/QALY)
3.7
84,300
25
160,500
$1,889
-$3,286
CostSaving ¶
Statin +
Ezetimibe**,††
26.1
140,500
71
303,300
$47,217
-$7,709
$130,000
Statin + PCSK9
inhibitor^^,††
27.5
335,300
33
680,800
$245,111
-$17,833
$334,000
Abbreviations: CV, cardiovascular; FH, familial hypercholesterolemia; ICER, incremental cost-effectiveness ratio; MACE, major
adverse cardiovascular event (nonfatal MI, nonfatal stroke, and cardiovascular death); NNT5, number-needed-to-treat; PCSK9,
proprotein convertase subtilisin/kexin type 9; QALY, quality-adjusted life year.
* In a scenario analysis, we evaluated the impact of “full treatment” in which all statin-tolerant patients who were not already
receiving statins were first treated with high-intensity statins, after which the entire FH subpopulation (n = 748,000 patients in
2015) was incrementally treated with ezetimibe or a PCSK9 inhibitor. The analytic horizon was lifetime (defined as when
patients reach 95 years of age). To reflect the precision of the model, person-years of treatment are rounded to the 100,000s;
MACE and QALYs are rounded to the 100s; costs are rounded to the millions; and ICERs to the 1000s.
† Number of patients that would need to be treated for 5 years to avert one MACE event.
⌃ All costs are reported in 2015 U.S. dollars. Future costs and QALYs are discounted 3% a year.
§ Patients deemed to be statin-intolerant (base-case prevalence = 10% of the FH population) received only ezetimibe.
|| The statin (full treatment) arm was compared with the statin (as treated) arm.
¶ Compared with the statin (as treated) arm, the statin (full treatment) arm costs $1.9 billion less and generates 160,500
additional QALYs. It is, therefore, an economically “dominating” option.
**Patients deemed to be statin-intolerant (base-case prevalence = 10%) received only ezetimibe.
†† The statin+ezetimibe and statin+PCSK9 inhibitor arms are compared with the statin (full treatment) arm.
^^ Patients deemed to be statin-intolerant (base-case prevalence = 10% of the FH population) received only a PCSK9 inhibitor.
©Institute for Clinical and Economic Review, 2015
Page 40
Secondary Prevention Among Patients with a Prior History of CVD and Intolerant of Statins
The base-case analysis modeled 1,460,000 statin-intolerant patients with a history of CVD. This
equated to approximately 85 million patient-years of treatment over the lifetime horizon. Key
results are reported in Table 16, and additional results are shown in Appendix 7 Table 7. Compared
with the control arm (no lipid-lowering therapy), treatment with PCSK9 inhibitors averted 1,254,400
MACE and produced 2,366,000 additional QALYs at an ICER of $274,000/QALY. As in the FH
population, ezetimibe’s clinical effects were less pronounced, but its incremental drug costs were
approximately 20% of those for PCSK9 inhibitors, resulting in an ICER of $145,000/QALY versus no
lipid-lowering therapy.
Table 16. Base-Case Clinical and Economic Outcomes Among Statin-Intolerant Patients with a
Prior History of CVD.*
Personyears of
treatment
(millions)
Total
MACE
averted
NNT5†
Control (no additional
QALYs
gained⌃
Incremental
Drug Costs⌃
(million $)
Incremental
Costs, Other CV
Care⌃
(million $)
ICER
($/QALY)
comparator
lipid-lowering therapy)
Ezetimibe§
85.0
446,100
56
847,000
$138,560
-$15,961
$145,000
PCSK9 inhibitor§
85.4
1,254,400
21
2,366,000
$693,450
-$44,627
$274,000
Abbreviations: CV, cardiovascular; ICER, incremental cost-effectiveness ratio; MACE, major adverse cardiovascular event
(nonfatal MI, nonfatal stroke, and cardiovascular death); NNT5, number-needed-to-treat; QALY, quality-adjusted life year.
* In the base case, we assumed that 10% of the population was statin-intolerant. Patients who had a prior history of
cardiovascular disease received incremental treatment with ezetimibe or a PCSK9 inhibitor (n = 1,460,000 in 2015). The analytic
horizon was lifetime (defined as until patients reached the age of 95 years). To reflect the precision of the model, person-years
of treatment are rounded to the 100,000s; MACE and QALYs are rounded to the nearest 100s; costs are rounded to the
millions; and ICERs to the 1000s.
† Number of patients that would need to be treated for 5 years to avert one MACE event.
⌃ All costs are reported in 2015 U.S. dollars. Future costs and QALYs are discounted 3% a year.
§ Both the ezetimibe and PSCK9 inhibitor arms are compared with the control (no additional lipid-lowering therapy) arm.
©Institute for Clinical and Economic Review, 2015
Page 41
One-Way Sensitivity Analyses:
In one-way sensitivity analyses, we examined the impact of varying individual input parameters
while holding all others constant (Figure 5). The ICER was very sensitive to changes in drug cost and
the analytic horizon, with higher drug costs and shorter analytic horizons worsening the costeffectiveness of the therapy.
Figure 5. One-way sensitivity analyses among statin-intolerant patients with a history of CVD.
Base-case ICER = $274,000/QALY. The ICER was very sensitive to changes in drug cost and the analytic horizon, with lower drug
costs and longer analytic horizons improving the cost-effectiveness of the therapy. Abbreviations: CVD, cardiovascular disease;
ICER, incremental-cost-effectiveness ratio; LDL-C, low-density lipoprotein cholesterol; PCSK9, proprotein convertase
subtilisin/kexin type 9; QALY, quality-adjusted life year.
* In this sensitivity analysis, we vary the effect of LDL-C reduction by a PCSK9 inhibitor relative to an equivalent reduction in
LDL-C by a statin. The base case assumes that a 1mmol/dL reduction in LD-LC by a PCSK9 inhibitor produces the same effect on
CVD outcomes as a 1mmol/dL reduction in LDL-C by a statin; this one-way analysis increases or decreases the relative effect of
PCSK9 inhibitors on CVD outcomes by 25%.
©Institute for Clinical and Economic Review, 2015
Page 42
Secondary Prevention Among Patients with a Prior History of CVD and LDL-C ≥ 70mg/dL on Statin
Therapy
The base-case analysis modeled 7,271,000 patients with a history of CVD and an LDL-C level ≥
70mg/dL despite statin therapy, equating to approximately 410 million patient-years of treatment
over the lifetime horizon. Key results are reported in Table 17, and additional results are shown in
Appendix 7 Table 8.
Compared with the control arm, treatment with PCSK9 inhibitors averted 5,621,800 MACE and
produced 10,573,800 additional QALYs at an ICER of $302,000/QALY.
Table 17. Base-Case Clinical and Economic Outcomes Among Patients with a Prior History of CVD
and LDL-C ≥ 70mg/dL on Statin Therapy.*
Personyears of
treatment
(millions)
Total
MACE
averted
NNT5
†
QALYs
gained⌃
Statin
Incremental
Drug Costs⌃
(million $)
Incremental
Costs, Other
CV Care⌃
(million $)
ICER
($/QALY)
comparator
Statin +
Ezetimibe§
409.1
2,253,800
51
4,345,900
$673,155
-$85,520
$135,000
Statin + PCSK9
inhibitor§
416.9
5,621,800
21
10,573,800
$3,406,692
-$210,702
$302,000
Abbreviations: CV, cardiovascular; ICER, incremental cost-effectiveness ratio; LDL, low-density lipoprotein; MACE, major adverse cardiovascular
event (nonfatal MI, nonfatal stroke, and cardiovascular death); NNT, number-needed-to-treat; PCSK9, proprotein convertase subtilisin/kexin
type 9; QALY, quality-adjusted life year.
* In the base case, patients with pre-existing CVD and LDL-C ≥ 70mg/dL on statin therapy received incremental therapy with ezetimibe or a
PCSK9 inhibitor (n = 7,271,000 in 2015). The analytic horizon was lifetime (defined as until patients reached the age of 95 years). To reflect
the precision of the model, person-years of treatment are rounded to the 100,000s; MACE and QALYs are rounded to the 100s; costs are
rounded to the millions; and ICERs to the 1000s.
† Number of patients that would need to be treated for 5 years to avert one MACE event.
⌃ All costs are reported in 2015 U.S. dollars. Future costs and QALYs are discounted 3% a year.
§ Both the statin+ezetimibe and statin+PSCK9 inhibitor arms are compared with the statin-only arm.
©Institute for Clinical and Economic Review, 2015
Page 43
One-Way Sensitivity Analyses:
In the one-way sensitivity analyses, we examined the impact of varying individual input parameters
while holding all others constant (Figure 6). As with the sensitivity analyses reported above, the
ICER for adding PCSK9 therapy to background statin therapy was very sensitive to changes in drug
cost and the analytic horizon, with higher drug costs and shorter analytic horizons worsening the
cost-effectiveness of the therapy.
Figure 6. One-way sensitivity analyses among patients with cardiovascular disease but LDL-C≥
70mg/dL on statin therapy.
Base-case ICER = $302,000/QALY. The ICER was very sensitive to changes in drug cost and the analytic horizon, with lower drug
costs and longer analytic horizons improving the cost-effectiveness of the therapy. Abbreviations: CVD, cardiovascular disease;
ICER, incremental-cost-effectiveness ratio; LDL-C, low-density lipoprotein cholesterol; PCSK9, proprotein convertase
subtilisin/kexin type 9; QALY, quality-adjusted life year.
* In this sensitivity analysis, we vary the effect of LDL-C reduction by a PCSK9 inhibitor relative to an equivalent reduction in
LDL-C by a statin. The base case assumes that a 1mmol/dL reduction in LDL-C by a PCSK9 inhibitor produces the same effect on
CVD outcomes as a 1mmol/dL reduction in LDL-C by a statin; this one-way analysis increases or decreases the relative effect of
PCSK9 inhibitors on CVD outcomes by 25%.
©Institute for Clinical and Economic Review, 2015
Page 44
Scenario Analysis: Initiating Therapy with Ezetimibe or PCSK9 After Incident (First-Ever) MI
In a scenario analysis, we explored the effect of only initiating therapy immediately after an incident
MI. All patients who had an incident, first-ever MI in 2015 who were receiving statin therapy, if able
to tolerate it, received ezetimibe or a PCSK9 inhibitor (n =169,000). Results are shown in Table 18
below. ICERs are lower than in the base case analysis for all secondary prevention ($170,000/QALY
and $74,000/QALY for PCSK9s and ezetimibe respectively) due to a greater reduction in the
absolute number of MACE events. NNT5 estimates for PCSK9 inhibitors and ezetimibe were 15 and
37 in this analysis.
Table 18. Scenario Analyses: Clinical and Economic Outcomes Among Patients Initiating
Ezetimibe or a PCSK9 Inhibitor After Incident (First-Ever) MI.*
Personyears of
treatment
(millions)
Total
MACE
averted
NNT5†
QALYs
gained⌃
Statin§
Statin +
Ezetimibe||,¶
Statin + PCSK9
inhibitor**,¶
Incremental
Drug Costs⌃
(million $)
Incremental
Costs, Other
CV Care⌃
(million $)
ICER
($/QALY)
comparator
3.3
16,800
37
64,200
$5,844
-$1,075
$74,000
3.5
43,200
15
159,200
$29,751
-$2,692
$170,000
Abbreviations: CV, cardiovascular; ICER, incremental cost-effectiveness ratio; LDL-C, low-density lipoprotein; MACE, major
adverse cardiovascular event (nonfatal MI, nonfatal stroke, and cardiovascular death); NNT5, number-needed-to-treat; PCSK9,
proprotein convertase subtilisin/kexin type 9; QALY, quality-adjusted life year.
* In this scenario analysis, all patients who had an incident (first-ever) MI in 2015, were receiving statin therapy if tolerated,
received incremental treatment with ezetimibe or a PCSK9 inhibitor (n = 169,000). Ten percent of the population was assumed
to be statin-intolerant. The analytic horizon was lifetime (defined as when patients reach the age of 95 years). To reflect the
precision of the model, person-years of treatment are rounded to the 100,000s; MACE and QALYs are rounded to the 100s;
costs are rounded to the millions; and ICERs to the 1000s.
† Number of patients that would need to be treated for 5 years to avert one MACE event.
⌃ All costs are reported in 2015 U.S. dollars. Future costs and QALYs are discounted 3% a year.
§ Patients deemed to be statin-intolerant (base-case prevalence = 10%) received no lipid-lowering therapy.
|| Patients deemed to be statin-intolerant (base-case prevalence = 10%) received only ezetimibe.
¶ Both the statin+ezetimibe and statin+PSCK9 inhibitor arms are compared with the statin-only arm.
** Patients deemed to be statin-intolerant (base-case prevalence = 10%) received only a PCSK9 inhibitor.
Threshold Analyses:
We evaluated the drug costs at which PCSK9 inhibitors would be considered cost-effective under
conventional willingness-to-pay thresholds of $50,000/QALY, $100,000/QALY, and $150,000/QALY.
For these analyses, we only considered total cardiovascular costs in the incremental costeffectiveness ratio, employed a lifetime analytic horizon and discounted future costs at 3% a year.
Results are presented in Table 19 on the following page.
©Institute for Clinical and Economic Review, 2015
Page 45
When weighted by the size of the three major subpopulations of interest (i.e., FH, CVD statinintolerant, CVD not at LDL-C target), threshold prices were $3,166, $5,404, and $7,735 for $50,000,
$100,000, and $150,000 per QALY respectively.
Table 19. Threshold analyses: Annual drug cost at which PCSK9 inhibitors would be cost-effective
in high-risk subpopulations under varying willingness-to-pay thresholds.*
Patient Subpopulation
FH on statin (as treated) + statinintolerant †
FH will full treatment with statin or
statin-intolerant §
Pre-existing CVD, LDL-C ≥ 70 mg/dL,
and statin-intolerant ||
Pre-existing CVD, LDL-C ≥ 70 mg/dL on
current statin regimen ¶
After first-ever MI
ALL SUBPOPULATIONS
$50,000/QALY
WTP threshold
$100,000/QALY
$150,000/QALY
$3,400
$5,700
$8,000
$3,000
$5,000
$7,000
$3,400
$5,800
$8,300
$3,100
$5,300
$7,600
$4,300
$3,166
$7,600
$5,404
$10,800
$7,735
Abbreviations: CVD, cardiovascular disease; FH, familial hypercholesterolemia; LDL, low-density lipoprotein; PCSK9,
proprotein convertase subtilisin/kexin type 9; QALY, quality-adjusted life year; WTP, willingness-to-pay.
* Only drug costs and costs related to cardiovascular care were included in the ICER for these analyses. The analytic horizon
was lifetime (defined as until patients reached 95 years of age), and future costs and QALYs were discounted at 3% a year. To
reflect precision in the model, the reported threshold drug costs are rounded to the nearest 100s.
† Patients who met the operational definition of FH and are either already receiving statin therapy or deemed statinintolerant (10% of the population) received incremental therapy with a PCSK9 inhibitor (n = 600,000 in 2015). Complete
results of this analysis are presented in Table 14 above.
§ All statin-tolerant patients who met the operational definition of FH but were not already receiving statins were first
treated with high-intensity statins, after which the entire FH subpopulation was incrementally treated with a PCSK9 inhibitor
(n = 748,000 patients in 2015). Complete results of this analysis are presented in Table 15 above.
|| Ten percent of the population was assumed to be statin-intolerant (n = 1,460,000 in 2015). Complete results of this
analysis are presented in Table 16 above.
¶ Patients with pre-existing CVD and LDL-C ≥ 70mg/dL already receiving statin therapy received incremental therapy with a
PCSK9 inhibitor (n = 7,271,000 in 2015). Complete results of this analysis are presented in Table 17 above.
** Patients who had an incident (first-ever) MI in 2015 and were receiving statin therapy if able to tolerate it received
incremental therapy with a PCSK9 inhibitor (n = 169,000). Complete results of this analysis are presented in Table 18 above.
©Institute for Clinical and Economic Review, 2015
Page 46
6.3 Health System Value
In addition to the incremental cost-effectiveness of PCSK9 inhibitors, we sought to estimate the
total budgetary impact of PCSK9 inhibitors in each of the three target populations defined above.
Assessment of budget impact as well as comparison to defined thresholds for policy intervention
and presentation of “value-based price benchmarks” are described in further detail below.
Budget Impact Model: Methods
We used the same model employed for the care value analysis to estimate total budgetary impact.
Budgetary impact was defined as the total incremental cost of the therapy in each population which
equals the incremental health care costs (including drug costs) minus any offsets in these costs from
averted cardiovascular events. All costs were undiscounted and estimated over one- and five-year
time horizons. The five-year timeframe was of primary interest, given the potential for cost offsets
to accrue from averted cardiovascular events. In addition to FH and patients with a history of CVD
who are (a) statin intolerant or (b) not at LDL-C target on statin therapy, we also considered the
budgetary impact if the treated population were limited to the higher-risk subset of patients with a
history of CVD who received PCSK9 inhibitors immediately following an incident (i.e., first-ever) MI
in 2015.
ICER’s methods for estimating budget impact and calculating benchmark prices are described in
detail at (http://www.icer-review.org/impact-and-outcomes/value-assessment-project/). Briefly,
our calculations assume that utilization of new drugs is “unmanaged” – i.e., without payer or
pharmacy benefit management controls in place – to provide an upper bound for likely patterns of
drug uptake by five years after launch. We examine six characteristics of the drug and marketplace
to estimate unmanaged drug uptake. These characteristics are listed below:

Magnitude of improvement in clinical safety and/or effectiveness

Patient-level burden of illness

Patient preference (ease of administration)

Proportion of eligible patients currently being treated

Primary care vs. specialty clinician prescribing/use

Presence or emergence of competing treatments of equal or superior effectiveness
Based on our assessment of these criteria, we assign a new drug to one of four categories of
unmanaged drug uptake patterns: 1) very high (75% uptake by year 5); 2) high (50% uptake by year
5); 3) intermediate (25% uptake by year 5); and 4) low (10% uptake by year 5).
©Institute for Clinical and Economic Review, 2015
Page 47
For patients with FH, we assumed a “very high” uptake pattern (75% at five years) for the two
drugs, combined, driven largely by the perceived acuity of the need for treatment of this
population. We assigned an “intermediate” uptake pattern (25% at five years) for the two drugs,
combined, in the secondary prevention population, given the availability of two new agents of
comparable effectiveness, as well as alternative treatments. For consistency, uptake was assumed
to occur in equal proportions across the five-year timeframe, and we adjusted both drug costs and
cost offsets accordingly. For example, in populations estimated to have a 25% 5-year uptake, 5% of
patients would be assumed to initiate therapy each year. Patients initiating therapy in year 1 would
accrue all drug costs and cost offsets over the full five years, but those initiating in other years
would only accrue a proportional amount of 5-year costs.
Using this approach to estimate potential budget impact, we then compared our estimates to a
budget impact threshold that represents a potential trigger for policy mechanisms to improve
affordability through changes to pricing, payment, or patient eligibility. As described in ICER’s
methods presentation (http://www.icer-review.org/wp-content/uploads/2014/01/Slides-on-valueframework-for-national-webinar1.pdf), this threshold is based on an underlying assumption that
health care costs should not grow much faster than growth in the overall national economy. From
this foundational assumption, our potential budget impact threshold is derived using an estimate of
growth in US gross domestic product (GDP) +1%, the average number of new molecular entity
approvals by the FDA each year, and the contribution of spending on retail and facility-based drugs
to total health care spending. Calculations can be found in Table 20 below.
Table 20. Calculation of Potential Budget Impact Threshold.
Item
1
2
3
4
5
6
7
8
Parameter
Growth in US GDP, 2015-2016 (est.) +1%
Total health care spending ($)
Contribution of drug spending to total health
care spending (%)
Contribution of drug spending to total health
care spending ($) (Row 2 x Row 3)
Annual threshold for net health care cost
growth for ALL new drugs (Row 1 x Row 4)
Average annual number of new molecular
entity approvals, 2013-2014
Annual threshold for average cost growth per
individual new molecular entity
(Row 5 ÷ Row 6)
Annual threshold for estimated potential
budget impact for each individual new
molecular entity (doubling of Row 7)
©Institute for Clinical and Economic Review, 2015
Estimate
3.75%
$3.08 trillion
13.3%
$410 billion
Source
World Bank, 2015
CMS NHE, 2014
CMS NHE, Altarum Institute,
2014
Calculation
$15.4 billion
Calculation
34
FDA, 2014
$452 million
Calculation
$904 million
Calculation
Page 48
For 2015-16, therefore, the five-year annualized potential budget impact threshold that should
trigger policy actions to manage affordability is calculated to total approximately $904 million per
year. In this report, each PCSK9 inhibitor is considered as an individual new drug, so the budget
impact threshold for each drug is $904 million and $1.8 billion for the two drugs combined.
Potential Budget Impact and the Value-based Price Benchmark
We combine consideration of the potential budget impact with the care value threshold prices
presented in Section 6 above to calculate a value-based price benchmark for each new drug. This
price benchmark begins with the care value price range to achieve cost-effectiveness ratios of
$100,000-$150,000 per QALY for the population being considered but has an upper limit
determined by the price at which the new drug would exceed the potential budget impact
threshold of $904 million. If the potential budget impact does not exceed $904 million, then the
value-based price benchmark remains the full care value price range.
Budget Impact Model: Results
The following tables present the budgetary impact of one year and five years of PCSK9 inhibitor
therapy in the populations studied, assuming the uptake patterns previously described. Detailed
information on cost offsets and clinical events averted can be found in the Appendix 7 Tables 10-15.
Findings based on the assumption that 75% of FH and 25% of CVD patients would initiate PCSK9
therapy by five years (i.e., 15% and 5% of these eligible populations each year), with drug costs and
cost offsets adjusted accordingly, are presented in Table 21 on the following page. Calculations for
adjustment of drug costs and cost offsets are provided in Appendix 7 Table 16. Results are
presented for both one-year and five-year time horizons.
Results from the budget impact model showed that if both the FH and CVD populations are treated
with the uptake pattern assumptions mentioned above, 527,000 individuals would receive PCSK9
therapy in the first year. After one year of PCSK9 treatment, cost offsets due to reduced
cardiovascular adverse events range from $592 per patient with FH to $1,010 per patient for
patients with CVD who are statin-intolerant. Including this cost offset, one-year budget impact is
still estimated to be quite high: approximately $7.2 billion for all patient populations.
Over the entire 5-year time horizon, we estimate that approximately 2.6 million persons would
receive PCSK9 inhibitor therapy for one or more years. Drug cost and cost-offset adjustments for
the full 5-year time horizon are described in detail in Appendix 7; across this timeframe the
weighted budgetary impact (i.e., adjusted for differing periods of drug utilization and associated
cost-offsets for different patients) ranges between $40,000 and $41,000 per patient for each
subpopulation. Total budgetary impact over five years is approximately $19 billion, $15 billion, and
$74 billion for the FH, CVD statin-intolerant, and CVD not at LDL-C target subpopulations. When
©Institute for Clinical and Economic Review, 2015
Page 49
these 5-year budget impact figures are annualized, they equal $21.4 billion in net health care cost
growth per year. This annualized potential budget impact is well above the budget impact threshold
of $1.8 billion for the two drugs combined. In order to not exceed this budget impact threshold,
approximately 1% of eligible patients could be treated at the average list price of $14,350 per year.
Table 21. Total Budgetary Impact of PCSK9 Inhibitors based on Assumed Patterns of Uptake, by
Subpopulation.
Analytic Horizon = 1 Year
Eligible
Population
(thousands)
FH
CVD, statinIntolerant
CVD, not at
LDL-C target
TOTAL
CVD, first-ever
MI only
Analytic Horizon = 5 Years
Number
Treated
(thousands)
Weighted BI
per Patient
($)*
Total BI,
$
(billions)
605
91
$13,824
$1.3B
1,460
73
$13,496
7,271
364
9,335
169
Number
Treated
(thousands)
Weighted BI per
Patient ($)*
Total BI
per yr, $
(billions)
453
$41,345
$3.7B
$1.0B
365
$40,446
$3.0B
$13,529
$4.9B
1,818
$40,562
$14.7B
527
$13,576
$7.2B
2,636
$40,681
$21.4B
9
$13,391
$0.1B
42
$39,931
$0.3B
NOTE: Subpopulation figures may not sum to total due to rounding
BI: Budget impact
*Weighted budget impact calculated by subtracting cost offsets from drug costs for one-year horizon. For five-year horizon,
drug costs and cost offsets apportioned assuming 20% of patients in uptake target initiate therapy each year. Those initiating in
year 1 receive full drug costs and cost offsets, those initiating in year 2 receive 80% of drug costs and cost offsets, etc.
Figure 7 on the following page provides findings of multiple analyses that give perspective on the
relationship between varying possible drug prices, cost-effectiveness ratios, drug uptake patterns,
and potential budget impact. The vertical axis shows the annualized budget impact, and the
horizontal axis represents the percentage of eligible patients treated over a five-year period. The
colored lines demonstrate how quickly the annual budget impact increases with increasing
percentages of patients treated at four different drug prices: those at which the cost/QALY =
$50,000, $100,000 and $150,000; and the list price used in this analysis ($14,350).
As can be seen in Figure 7, even at a drug cost of $3,166 dollars per year, the cost at which the
cost/QALY = $50,000, if 50% of all eligible patients are ultimately treated over a five-year time
period the annualized budget impact is approximately $5.6 billion per year. At the list price of
$14,350 used for this report, if only 25% of eligible patients receive treatment, the annualized
budget impact is nearly $19 billion, meaning that over the five-year period a total of almost $100
billion would have been added to health care costs in the United States.
©Institute for Clinical and Economic Review, 2015
Page 50
Figure 7. ICER value graph combining cost-effectiveness and potential budget impact analyses.
Colored lines represent the impact on annualized budget impact of different uptake patterns
(eligible patients treated) at the actual list price of the drug (dashed line) and at drug prices needed
to achieve common incremental cost-effectiveness ratios.
Estimated Cost/QALY: $296,850
($14,350 annual drug price)
$36
Annual Budget Impact (Billions)
$32
Cost/QALY $150,000
($7,735 annual drug
price)
$28
Cost/QALY $100,000
($5,404 annual drug
price)
$24
$20
$16
Cost/QALY $50,000
($3,166 annual drug
price)
$12
$8
$4
$0
0%
25%
50%
75%
100%
% Eligible Patients Treated
6.4 Draft Value-based Benchmark Prices
Our draft value-based benchmark prices for each key subpopulation and for the overall treated
population are provided in Table 22 on the following page. As noted in the ICER methods document
(http://www.icer-review.org/wp-content/uploads/2014/01/Slides-on-value-framework-fornational-webinar1.pdf), the draft value-based benchmark price for a drug is defined as the care
value price range that would achieve cost-effectiveness ratios between $100,000 and $150,000 per
QALY gained, limited, if the results require, by the price at which the $904 million budgetary impact
threshold would be exceeded. Detailed calculations for the value-based price benchmarks
presented below are available in Appendix 7 Table 17.
As shown in the table on the following page, if only the FH or the CVD statin-intolerant populations
were treated, the entire care value price range is lower than the price at which the potential budget
impact threshold would be exceeded. Thus, the value-based price benchmark for these two
subpopulations is the care value price range. This is not surprising given the relatively small size of
each of these populations. In contrast, the care value price range for the much larger population of
©Institute for Clinical and Economic Review, 2015
Page 51
patients with CVD not at LDL-C target is higher than the maximum price that would not exceed the
budget impact threshold.
When all subpopulations are combined, the care value price range is $5,404-$7,735. But this price
range is higher than the maximum price that could be charged before exceeding the potential
budget impact threshold ($2,177). Therefore, the draft ICER value-based price benchmark for each
of the new PCSK9 drugs, with all the assumptions mentioned previously regarding 5-year uptake
patterns and cost offsets, is $2,177. This figure represents an 85% discount from the full wholesale
acquisition cost assumed in our analysis ($14,350).
Table 22. Draft value-based price benchmarks for PCSK9 inhibitor therapy.
Population
FH
(n=453,443)
CVD statin-intolerant
(n=364,948)
CVD not at LDL target
(n=1,817,788)
Care Value Price:
$100K/QALY
Care Value Price:
$150K/QALY
Draft Value-Based
Price Benchmark
$8,000
Max Price at
Potential Budget
Impact Threshold
$10,278
$5,700
$5,800
$8,300
$12,896
$5,800-$8,300
$5,300
$7,600
$2,976
$2,976
$5,700-$8,000
TOTAL (n=2,636,179)
$5,404
$7,735
$2,177
$2,177
Abbreviations: FH: familial hypercholesterolemia; CVD: cardiovascular disease; LDL: low-density lipoprotein; QALY:
quality-adjusted life year
6.5 Summary and Comment
The results of our cost-effectiveness analysis suggest that the use of PCSK9 inhibitors may produce
substantial reductions in non-fatal MIs, non-fatal strokes, and cardiovascular deaths over the
lifetime analytic horizon. The NNT5 (number of patients that would be needed to be treated for 5
years to avoid one major adverse cardiovascular event) of 28 for PCSK9 inhibitors appears to be
relatively low; despite this, however, treatment with PCSK9 inhibitors generates cost-effectiveness
ratios that far exceed commonly-accepted thresholds such as $100,000/QALY.79 Achieving costeffectiveness at a threshold of $100,000/QALY would require price reductions of 60% to 63%
compared with current prices. And the results of our analysis of potential budget impact suggest
that even deeper reductions may be required to avoid excessive cost burdens to the health care
system. Our value-based price benchmark for each PCSK9 inhibitor is $2,177 annually, which
represents an 85% reduction from the list price of $14,350.
The high costs of specialty drugs – including novel chemotherapeutic agents for cancer and
treatments for Hepatitis C – have generated considerable debate about their societal value.125-127
The opportunity cost of these high drug prices is a function of the incidence and prevalence of the
©Institute for Clinical and Economic Review, 2015
Page 52
targeted disease, as well as the indicated duration of treatment.127 An expensive therapy for a rare
disease that affects a small number of patients may only have a small impact on the health care
budget. On the other hand, PCSK9 inhibitors are meant to be lifelong therapy for a large and
growing population with pre-existing CVD, and their high price may have a sizeable effect on total
health care spending.
The key strengths of this analysis are the modeling of a nationally representative cohort, long-term
follow-up, and real world cost estimates. Extensive one-way and scenario-based sensitivity analyses
were used to examine the impact of uncertainty about the input parameters on the results. We
found that across all subpopulations, results were most sensitive to changes in the price of PCSK9
inhibitors and the length of the time horizon. However, in none of the one-way sensitivity analyses
(other than those on price) did the ICERs for PCSK9 inhibitor therapy fall below $219,000 per
QALY. As expected, PCSK9 inhibitors were most cost-effective in the patients at highest risk of
adverse events, i.e., patients after their first-ever MI. Future studies should examine the role of risk
stratification in improving the effectiveness and cost-effectiveness of PCSK9 inhibitors.
Our analysis has several key limitations that merit attention. First, there are no long-term
effectiveness data for PCSK9 inhibitors, although short-term studies suggest that they lower the risk
of MI and cardiovascular death. While modeling the lifetime analytic horizon as we have done in
this analysis captures all potential benefits and costs related to the intervention based on our
current knowledge, extrapolating to several decades the results of trials that are 3-6 months long
increases the uncertainty in the results. For instance, if short-term LDL-C reductions observed in the
trials are not sustained, we will have over-estimated CVD risk reduction benefits. There is even
greater uncertainty about their impact on stroke. In line with the “LDL hypothesis,” our base case
assumed that a 1 mmol/L reduction in LDL-C levels would produce an identical reduction in MI,
stroke, and cardiovascular death, irrespective of whether this reduction was achieved using a PCSK9
inhibitor or a statin. We varied the effect of PCSK9 inhibitors on cardiovascular event rates (from 25% to +25% relative to the base case), and found this to be a moderately sensitive parameter; in
the FH population, for example, cost-effectiveness ratios ranged from $250,000 to $359,000 per
QALY gained. Future long-term studies will address whether the beneficial effects of statins on
cardiovascular events can truly be extrapolated to PCSK9 inhibitors.
Second, given lack of evidence of clinically significant harms in short-term PCSK9 trials, we did not
model drug-related adverse events. Injection-site reactions were common in PCSK9 inhibitor trials,
but we assumed these are mild and transient and did not affect patients’ quality of life. Although
the incidence of these events was not significantly different among patients receiving PCSK9
inhibitors and patients receiving an injectable placebo used in clinical trials, injection-site reactions
may in fact represent a small increase in costs and disutility relative to the orally administered
comparators in this analysis. If future studies reveal more serious or more frequent adverse events,
or substantial disutility arising from the need for parenteral administration of PCSK9 inhibitors, we
©Institute for Clinical and Economic Review, 2015
Page 53
will have overestimated the cost-effectiveness of PCSK9 inhibitors, and this analysis would need to
be updated to account for this new safety information.
Third, this analysis was performed assuming that patients continued taking the medication strategy
assigned at baseline for years afterward, and did not directly account for decrements in medication
adherence, stopping the assigned medication strategy, or switching to another strategy over time.
Our model partially accounts for medication non-adherence to the extent that its effect on clinical
efficacy is already captured in the observed risk reduction seen in clinical trials and our populationbased approach accounted for background statin therapy in the population. However, real-world
adherence may vary from that observed in clinical trials based on age, educational status, comorbidities, and cost-sharing.128 It is plausible that adherence to PCSK9 inhibitors – which have to
be self-injected and may have higher co-pays – may be lower in the real world compared with that
observed in clinical trial populations. At the same time, a monthly or biweekly regimen may
improve adherence over a once-daily dose. Future studies must examine the impact of nonadherence on effectiveness and safety.
Fourth, our model incorporated the entire cohort of US adults aged 35 to 74 years in the year 2015
and followed them until the age of 95 years. Treatment of FH often begins in childhood or
adolescence with premature coronary disease often manifesting in the third decade of life; we
therefore may not have captured the entire clinical and economic burden of FH in the population or
the benefits of LDL-C lowering in childhood or young adulthood. However, there are currently no
data about the efficacy or safety of PCSK9 inhibitors in children, and until such data become
available, any assumptions about the benefits and risks of PCSK9 inhibitors in children would be
purely speculative. We assumed that the elevated cardiovascular risk among patients with FH is
entirely mediated by their high serum levels of LDL-C along with characteristics measured in the
NHANES risk factor survey. If an FH patient with a certain LDL cholesterol level is at a higher risk for
a cardiovascular event compared with an otherwise identical patient with the same LDL-C level but
without FH (perhaps due to onset of exposure to high LDL-C earlier in life, leading to accelerated
atherosclerosis), we will have underestimated the clinical and economic burden of FH, and possibly
the cost-effectiveness of PCSK9 inhibitors in this population. On the other hand, it is likely that our
operational definition of FH (with LDL-C level ≥250mg/dL off statins or ≥200mg/dL on statin
therapy) identifies a subgroup of patients at a relatively higher risk of events given that the clinical
diagnosis of FH can be made in adults with an LDL-C level greater than or equal to 190mg/dL who
have a family history of FH. We chose our operational definition to increase specificity in the
absence of family history data in the model and to approximate the previously published population
prevalence of FH (1 in 500). But the higher LDL-C threshold may have produced a study cohort at
higher-than-average risk of cardiovascular events. Applying a lower LDL-C threshold to define FH
(e.g., LDL-C ≥ 190mg/dL) would substantially increase the number of individuals eligible for
additional lipid-lowering therapy and hence amplify the budget impact. At the same time, lowering
©Institute for Clinical and Economic Review, 2015
Page 54
the LDL-C threshold would lower the average risk of cardiovascular events in the eligible population,
thus reducing the cost-effectiveness of PCSK9 inhibitors in this subpopulation.
Fifth, our model assumes a constant drug cost over the analytic horizon, but long-term changes in
drug prices may be impacted by patent expiration, availability of generics or biosimilars, or
development of novel therapies. If drug-prices decrease over time, we will have underestimated
long-term cost-effectiveness and over-estimated budget impact.
Finally, our assumed levels of PCSK9 inhibitor uptake in the marketplace by five years were based
on reasoned assumptions, but actual uptake may vary from these estimates. We also present
potential budget impact across a range of uptake possibilities in sensitivity analyses.
©Institute for Clinical and Economic Review, 2015
Page 55
7. Summary of the Votes and Considerations for
Policy
7.1 About the New England CEPAC Process
During public meetings of the New England CEPAC, the Council deliberates and votes on key
questions related to the systematic review of the clinical evidence, an economic analysis of the
applications of the medical technologies or treatments under examination, and the supplementary
information presented. Council members are selected for three year terms and are intentionally
selected to represent a range of expertise and diverse perspectives. To maintain the objectivity of
New England CEPAC and ground the conversation in the interpretation of the published evidence,
members are not pre-selected based on the topic being addressed. Acknowledging that any
judgment of evidence is strengthened by real-life clinical and patient perspectives, clinical
representatives with expertise in the subject matter are recruited for each meeting topic and
provide input to Council members before the meeting to help clarify their understanding of the
interventions being analyzed in the evidence review. The same clinical experts serve as a resource
to the Council during their deliberation, and they help form recommendations with the Council on
ways the evidence can be applied to policy and practice.
At each meeting, after the Council votes, a Policy Roundtable discussion is held with the Council,
clinical experts, and representatives from provider groups, payers, and patient groups. This is
intended to bring stakeholders into the discussion on how best to apply the evidence to guide
patient education, clinical practice, and coverage and public policies. Participants on Policy
Roundtables are selected for their expertise on the specific meeting topic, are different for each
meeting, and do not vote on any questions.
At the October 27, 2015 meeting, the Council discussed issues regarding the application of the
available evidence to help patients, providers, and payers address the important questions related
to the management of high cholesterol. Following an evidence presentation and public comments,
the Council voted on key questions concerning the clinical effectiveness and value of PCSK9
inhibitors alirocumab and evolocumab. These questions are developed by the ICER research team
for each assessment, with input from the New England CEPAC Advisory Board to ensure that the
questions are framed to address the issues that are most important in applying the evidence to
support clinical practice and medical policy decisions. The voting results are presented in the
section below, along with comments reflecting considerations mentioned by the Council members
during the voting process.
©Institute for Clinical and Economic Review, 2015
Page 56
In its deliberations and voting related to value, the Council made use of a value assessment
framework with four different components of care value, a concept which represents the long-term
perspective, at the individual patient level, on patient benefits and the incremental costs to achieve
those benefits. The four components of care value are comparative clinical effectiveness,
incremental cost per outcomes achieved, additional benefits or disadvantages, and contextual
considerations regarding the illness or therapy.
Once they made an overall assessment of care value as low, intermediate, or high considering these
four components, the New England CEPAC then explicitly considered the affordability of PCSK9
inhibitors in assessing provisional health system value as low, intermediate, or high (see Figure 8
below and Figure 9 on the next page, as well as the detailed explanation that follows).
Figure 8. Care Value Framework
There are four elements to consider when deliberating on care value:
1. Comparative clinical effectiveness is a judgment of the overall difference in clinical outcomes
between two interventions (or between an intervention and placebo), tempered by the level of
certainty possible given the strengths and weaknesses of the body of evidence. The Council uses
the ICER Evidence Rating Matrix as its conceptual framework for considering comparative
clinical effectiveness.
2. Incremental cost per outcomes achieved is the average per-patient incremental cost of one
intervention compared to another to achieve a desired “health gain,” such as an additional
stroke prevented, case of cancer diagnosed, or gain of a year of life. Alternative interventions
are compared in terms of cost per unit of effectiveness, and the resulting comparison is
presented as a ratio: a “cost per outcome achieved.” Relative certainty in the cost and outcome
estimates continues to be a consideration. As a measure of incremental costs per outcomes
achieved, ICER follows common academic and World Health Organization (WHO) standards by
using cost per quality-adjusted life years (QALYs) and adopting thresholds at $100,000 per QALY
and $150,000 per QALY as guides to reasonable ratios of incremental costs per outcomes
achieved.
©Institute for Clinical and Economic Review, 2015
Page 57
3. Other benefits or disadvantages refers to any significant benefits or disadvantages offered by
the intervention to the individual patient, caregivers, the delivery system, other patients, or the
public that would not have been considered as part of the evidence on comparative clinical
effectiveness. Examples of additional benefits include mechanisms of treatment delivery that
require many fewer visits to the clinician’s office, treatments that reduce disparities across
various patient groups, and new potential mechanisms of action for treating clinical conditions
that have demonstrated low rates of response to currently available therapies. Additional
disadvantages could include increased burden of treatment on patients or their caregivers. For
each intervention evaluated, it will be open to discussion whether additional benefits or
disadvantages such as these are important enough to factor into the overall judgment of care
value. There is no quantitative measure for additional benefits or disadvantages.
4. Contextual considerations include ethical, legal, or other issues (but not cost) that influence the
relative priority of illnesses and interventions. Examples of contextual considerations include
whether there are currently any existing treatments for the condition, whether the condition
severely affects quality of life or not, and whether the condition affects priority populations.
There is no quantitative measure for the role of contextual considerations in an overall
judgment of care value.
In assessing provisional health system value, the Council was asked to vote whether interventions
represent a “high,” “intermediate,” or “low” value.
Figure 9. Health System Value Framework
1. Potential Health System Budget Impact is the estimated net change in total health care costs
over a 5-year time-frame.
2. Provisional “Health System Value” represents a judgment integrating consideration of the longterm care value of a new intervention with an analysis of its potential short-term budget impact
if utilization is unmanaged. The Council votes reflect a judgement on the provisional health
system value of an intervention.
3. Mechanisms to Maximize Health System Value is an action step, ideally supported by enhanced
early dialogue among manufacturers, payers, and other stakeholders.
©Institute for Clinical and Economic Review, 2015
Page 58
4. Achieved Health System Value is the real-world result of health care stakeholder efforts to
maximize the value of a given intervention.
Usually, the care value and the provisional health care system value of an intervention or approach
to care will align, whether it is “high,” “intermediate,” or “low.” For example, a treatment that is
judged to represent high care value from the perspective of per-patient costs and benefits will
almost always represent a high health system value as well. But health system value also takes into
consideration the short-term effects of the potential budget impact of a change in care across the
entire population of patients. Rarely, when the additional per-patient costs for a new care option
are multiplied by the number of potential patients treated, the short-term budget impact of a new
intervention of intermediate or even high care value could be so substantial that the intervention
would be “unaffordable” unless the health system severely restricts its use, delays or cancels other
valuable care programs, or undermines access to affordable health insurance for all patients by
sharply increasing health care premiums. Under these circumstances, unmanaged change to a new
care option could cause significant harm across the entire health system, in the short-term possibly
even outweighing the good provided by use of the new care option itself.
Provisional health system value builds upon the judgment of care value by integrating consideration
of the potential short-term budget impact of a new intervention, a figure highly dependent upon an
estimation of the potential uptake of the new drug across the entire population. In the ICER
framework, the theoretical basis for the budget impact threshold is based on societal willingness to
pay. This foundation rests upon the assumption that society would prefer health care costs to grow
at a rate that does not exceed growth in the overall national economy. ICER has used estimates
based on data from the World Bank, the Centers for Medicare & Medicaid Services (CMS), and
other public sources to calculate a budget impact threshold for individual new drugs or devices that
would identify those whose potential budget impact would contribute significantly to excessive
health care cost growth.
It should be noted that if, after considering potential budget impact, a health intervention judged to
have high care value receives a judgment of “low” provisional health system value from the Council,
this does not imply that the health system should not adopt the intervention; rather, the vote
indicates that policy makers should consider implementing mechanisms related to patient selection,
step therapy, pricing, and/or financing to ensure that the short-term budget impact of a high care
value intervention does not lead to more harm than good. New England CEPAC votes on provisional
health system value will therefore serve an important function by highlighting situations when
policymakers need to take action and work together to align care value with health system value.
©Institute for Clinical and Economic Review, 2015
Page 59
7.2 Comparative Clinical Effectiveness Voting Results
1. Is the evidence adequate to distinguish between the overall net health benefits of the
PCSK9 inhibitors Praluent® and Repatha™, excluding use in homozygous familial
hypercholesterolemia for which only Repatha has an indication?
Council Vote: 0 Yes (0%) 12 No (100%)
Sub populations include:
 Individuals with heterozygous familial hypercholesterolemia (HeFH) who are not at goal (LDL
<160mg/dL)
 Individuals with a history of atherosclerotic cardiovascular disease who cannot take statins or
who take statins but are not at goal (LDL < 70mg/dL)
For individuals with heterozygous familial hypercholesterolemia (HeFH) who are statin intolerant
or who take statins but are not at goal (<160mg/dL):
2. Is the evidence adequate to demonstrate that adding PCSK9 inhibitors to treatment
improves net health benefits?
Council Vote: 7 Yes (58%) 5 No (42%)
Comments: Council members voting yes noted that their votes assumed that lower LDL-C is
a sufficient indicator for improved outcomes. It was also noted that the higher risk in FH
populations was an important consideration and that this vote is based only on evidence
available at present. Members voting no suggested that while the early evidence looks
promising, more data is needed before conclusions can be drawn. They also expressed
concerns that lower LDL-C may not be a sufficient endpoint to indicate improved outcomes.
For individuals with a history of atherosclerotic cardiovascular disease who are statin intolerant:
3. Is the evidence adequate to demonstrate that adding PCSK9 inhibitors to treatment improves
net health benefits?
Council Vote: 4 Yes (33%) 8 No (67%)
Comments: CEPAC members voting no pointed primarily to the uncertainty around the LDL
hypothesis (whether a lower LDL-C leads to improved outcomes) given that this population may
not have risks as high as patients who have FH and therefore high cholesterol from birth.
Uncertainty was also discussed surrounding possible harms of the drugs given that LDL-C levels
as low as those produced by these drugs have not been seen before and may have unknown
long-term consequences.
©Institute for Clinical and Economic Review, 2015
Page 60
For individuals with a history of atherosclerotic cardiovascular disease who take statins but are
not at goal (LDL < 70mg/dL):
4. Is the evidence adequate to demonstrate that adding PCSK9 inhibitors to treatment improves
net health benefits?
Council Vote: 3 Yes (25%) 9 No (75%)
Comments: Similar to the previous question, Council members voting no cited a lack of
evidence that lowering LDL-C below 70mg/dL ultimately improves outcomes. Without this
evidence, many Council members felt unable to conclusively vote that PCSK9 inhibitors improve
net health benefit for this population. Council members voting yes noted the LDL is a central
mechanism of disease, and efforts should be taken to lower it when possible.
7.3 Care Value Voting Results
For individuals with heterozygous familial hypercholesterolemia (HeFH) who are statin intolerant
or who take statins but are not at goal (LDL <160mg/dL):
5. Given the available evidence, what is the care value of adding PCSK9 inhibitors vs. no additional
treatment?
Council Vote: 0 High (0%) 8 Intermediate (67%) 4 Low (33%)
Comments: A majority of the council found PCSK9 inhibitors to have an intermediate care value
for patients with HeFH. Members of the Council voting for an intermediate care value
acknowledged the high incremental cost-effectiveness ratios but were persuaded not to vote
“low” value because of the perceived increased risk and limited treatment options for patients
with FH in whom statins have not been sufficiently effective. It was suggested that PCSK9
inhibitors help to fill a critical unmet need for this population. Members voting for a low value
were persuaded by the high cost-effectiveness ratios in light of the residual uncertainty about
clinical benefits.
For individuals with a history of atherosclerotic cardiovascular disease who are statin intolerant:
6. Given the available evidence, what is the care value of adding PCSK9 inhibitors vs. no additional
treatment?
Council Vote: 0 High (0%) 5 Intermediate (42%) 7 Low (58%)
Comments: CEPAC found PCKS9 inhibitors to present a low to intermediate care value for this
population. Members felt that this population is not at as high of a risk as those with FH.
©Institute for Clinical and Economic Review, 2015
Page 61
For individuals with a history of atherosclerotic cardiovascular disease who take statins but are
not at goal (LDL < 70mg/dL):
7. Given the available evidence, what is the care value of adding PCSK9 inhibitors vs. no additional
treatment?
Council Vote: 0 High (0%) 2 Intermediate (17%) 10 Low (83%)
Comments: Rationale for a majority “low” vote was similar to that for the previous question;
council members also noted that this population has several alternative treatment options at
their disposal.
For the combined population of all patients in these groups
8. Given the available evidence, what is the care value of adding PCSK9 inhibitors vs. no additional
treatment?
Council Vote: 0 High (0%) 3 Intermediate (25%) 9 Low (75%)
Comments: Council members questioned the utility of lumping all patient populations together
into a single group; however, it was recognized that in real-world situations, policy makers may
not be able to tease out each sub-population.
7.4 Provisional Health System Value Voting Results
For individuals with heterozygous familial hypercholesterolemia (HeFH) who are statin intolerant
or who take statins but are not at goal (LDL <160mg/dL):
9. Given the available evidence, what is the provisional health system value of adding PCSK9
inhibitors vs. no additional treatment?
Council Vote: 0 High (100%) 2 Intermediate (17%) 10 Low (83%)
Comment: Council members voting for a low provisional health system value emphasized the
need to consider the high short-term costs from a societal perspective. Council members
emphasized that paying high prices for PCSK9 inhibitors will create the need for cuts in other
areas.
For Individuals with a history of atherosclerotic cardiovascular disease who are statin intolerant:
10. Given the available evidence, what is the provisional health system value of adding PCSK9
inhibitors vs. no additional treatment?
Council Vote: 0 High (100%) 0 Intermediate (0%) 12 Low (100%)
©Institute for Clinical and Economic Review, 2015
Page 62
For Individuals with a history of atherosclerotic cardiovascular disease who take statins but are
not at goal (LDL < 70mg/dL):
11. Given the available evidence, what is the provisional health system value of adding PCSK9
inhibitors vs. no additional treatment?
Council Vote: 0 High (100%) 0 Intermediate (0%) 12 Low (100%)
For the combined population of all patients in these groups
12. Given the available evidence, what is the provisional health system value of adding PCSK9
inhibitors vs. no additional treatment?
Council Vote: 0 High (100%) 0 Intermediate (0%) 12 Low (100%)
©Institute for Clinical and Economic Review, 2015
Page 63
7.5 Roundtable Discussion and Key Policy Recommendations
Following New England CEPAC’s deliberation on the evidence and subsequent voting, ICER
convened a Policy Roundtable intended to bring stakeholders into the discussion on how best to
apply the evidence to guide patient education, clinical practice, and coverage policies. The
Roundtable was composed of clinical experts, a patient, a representative from a national pharmacy
benefit management company, a regional insurer, a purchaser of state based health insurance, and
a representative from an academic medical center. Participants on Policy Roundtables are selected
for their expertise or experience related to the specific meeting topic, are different for each
meeting, and do not vote on any questions. The Policy Roundtable participants are displayed
below.
Table 23. Policy Roundtable Participants
Policy Roundtable Participants
Leslie Fish, PharmD
Vice President of Pharmacy, Fallon Health
Jonathan Karas
Patient Representative
Dolores Mitchell,
Executive Director, Group Insurance Commission
Patrick O’Gara, MD
Senior Physician, Brigham and Women’s Hospital
Professor of Medicine, Harvard Medical School
William Shrank, MD, MSHS
Senior Vice President, Chief Scientific Officer and Chief Medical Officer, Provider Innovation and Analytics,
CVS Health
Thomas Siepka, RPh, MS FACHE
Vice President, System Pharmacy and Outreach,
Dartmouth Hitchcock
Paul Thompson, MD
Chief of Cardiology, Hartford Hospital
Professor of Medicine, University of Connecticut
©Institute for Clinical and Economic Review, 2015
Page 64
The Roundtable discussion explored the implications of New England CEPAC’s votes for clinical
practice and medical policy, considered real life issues critical to developing best practice
recommendations in this area, and identified potential avenues for applying the evidence to
improve patient care. The main themes and recommended best practices from the conversation are
summarized below. The Policy Roundtable discussion reflected multiple perspectives and opinions;
therefore, none of the recommendations below should be taken as a consensus view held by all
participants.
1. Professional societies should take prompt action to update clinical guidance, including a likely
need to return to treatment goals based on target LDL-C levels.
Members of the Roundtable highlighted the need to link the dialogue surrounding PCSK9 inhibitors
to a broader discussion about clinical guidelines made by professional societies. Physicians rely on
clinical guidelines when making decisions about a particular diagnostic or therapeutic procedure.
Guidelines are intended to direct everyday clinical decision-making and reinforce the practice of
evidence based medicine. Roundtable participants noted that the emergence of PCSK9 inhibitors
has unveiled an urgent need for action from professional societies. As mentioned earlier in this
report, guidelines for the treatment of cholesterol have changed in recent years. In 2013, the
ACC/AHA released a guideline that moved away from recommending specific LDL-C levels as
treatment targets.5 While both the 2013 guideline and previous guidelines include strong
recommendations for use of statin therapy to treat individuals with cardiovascular disease, the
updated guideline does not recommend specific LDL-C targets. This guideline was developed when
little clinical evidence was available for treatment with PCSK9 inhibitors. Clinical participants on the
Roundtable voiced that LDL-C goals serve as a mechanism to measure patient progress.
Without specific treatment goals, clinicians may face challenges when prescribing statin therapy as
a preferred treatment. They may feel pressured to prescribe PCSK9 inhibitors to yield lower LDL-C
levels despite a lack of evidence supporting the hypothesis that lower LDL-C leads to improved
outcomes. An updated guideline that stratifies patients by their cardiovascular risk, includes
treatment goals based on target LDL-C levels, and titrates medication use to reach a specified target
would provide clinicians with a basis for discussion when examining treatment options with
patients. In order to remain relevant and effective within the context of a changed treatment
landscape, professional societies should produce updated guidelines that recommend specific LDL-C
targets based on patient characteristics.
©Institute for Clinical and Economic Review, 2015
Page 65
2. Payers should use prior authorization to enhance health system value by limiting treatment to
patients for whom extended trials of high-dose statins combined with ezetimibe have been
unsuccessful.
Due to the high cost of PCSK9 inhibitors, payers will likely administer utilization management
techniques to control aggregate health care costs. Specifically, prior authorization and step therapy
are used to stratify patients by risk and previous treatment in order to ensure that coverage policies
align with prudent clinical practice options. These coverage policies typically require patients to
first try a lower-cost, evidence based therapy to achieve a clinical goal. If a patient is unable to
achieve a goal with a specified therapy, a physician may escalate a patient to a more intensive and
costly option.
For high cost drugs, prior authorization criteria are generally based on an FDA indication, and
treatment is limited to patients for whom the evidence demonstrates greatest benefit. In addition
to the FDA indications, PCSK9 inhibitors will likely be limited to patients who have not reached
appropriate LDL-C after extended trials (likely 60-90 days) of high-dose statins combined with
ezetimibe. Clinical expert participants on the Roundtable felt that it was practical and reasonable
for patients to fail at least 2 statins, with one at the lowest dose, prior to use of PCSK9 inhibitors.
Statin non-adherence was also raised as a point of concern; one clinical expert expressed that
providers do not always know when a patient is taking statins as prescribed. Another participant
expressed that as part of coverage policies, payers may want to consider using results of lab tests to
demonstrate that 2 statins have been tried. If executed appropriately, these measures will enhance
the health system value of PCSK9 inhibitors. As noted earlier, unless the current treatment
guideline is updated, payers will likely face challenges when developing criteria and will be limited
in their ability to administer these tools. Discussion also turned to managing use of PCSK9 inhibitors
in key subpopulations as described below:
Patients with FH
There was consensus throughout the meeting that there is a significant unmet need among
patients with FH who are unable to lower cholesterol sufficiently using statins or ezetimibe.
The patient participant on the Roundtable shared his experience in dealing with FH as a
patient and as a parent of a child with FH, noting that some patients with FH must endure
intensive and expensive treatment like apheresis. Other Roundtable participants
mentioned that for those patients with clearly identified unmet need (e.g., homozygous FH
or very high LDLs on treatment) the discussed criteria could be relaxed. Roundtable
participants also stated that identifying patients with FH can be difficult. The clinical experts
acknowledged that LDL-C measures are most useful for understanding risk; specifically
noting that they would consider a patient to have clinical FH if they presented with an LDL >
190 mg/dL without any treatment. Measures such as genetic testing were deemed
unnecessary for most patients, as these tests can be costly and there are 1700 genetic
mutations that can cause FH.
©Institute for Clinical and Economic Review, 2015
Page 66
Primary Prevention and Off-Label Use
In the discussion of off-label use, there was general consensus that PCSK9 inhibitors should
not be used for primary prevention or for patients with comorbidities such as diabetes;
participants considered off-label use to be a slippery slope within the context of this
discussion.
3. Prior authorization criteria may need to require most patients who believe they are statin
intolerant to be re-tried on statins.
Many patients report adverse effects associated with statin therapy; however, the field lacks
consensus on a definition of statin intolerance. Based on feedback from the clinical expert
participants and a review of existing definitions, it may be considered reasonable by some to
operationalize the recent National Lipid Association (NLA) definition of statin intolerance when
making treatment decisions. In June 2014 National Lipid Association Expert Panel on Statin
Intolerance suggested the following definition for statin intolerance:
Statin intolerance is a clinical syndrome characterized by the inability to tolerate at least 2
statins: one statin at the lowest starting daily dose AND another statin at any daily dose, due to
either objectionable symptoms (real or perceived) or abnormal lab determinations, which are
temporally related to statin treatment and reversible upon statin discontinuation, but
reproducible by re-challenge with other known determinants being excluded (such as
hypothyroidism, interacting drugs, concurrent illnesses, significant changes in physical activity
or exercise, and underlying muscle disease). Specifically, the lowest starting statin daily dose, is
defined as rosuvastatin 5 mg, atorvastatin 10 mg, simvastatin 10 mg, lovastatin 20 mg,
pravastatin 40 mg, fluvastatin 40 mg, and pitavastatin 2 mg1.
This definition may prove useful in clinical practice, but it is important to note that there are
limitations to understanding statin intolerance. The prevalence of true statin intolerance is
uncertain and ‘perceived symptoms’ are not always related to statin therapy. One clinical expert
participant shared that many people who think they have statin intolerance actually can tolerate
the medications. Two studies referenced in this report examined statin intolerance; one found a
10% incidence of mild to moderate muscle symptoms for patients on high intensity statin therapy 17
and the second a 9.4% incidence of muscle symptoms in statin-naïve patients treated with
atorvastatin 80 mg daily compared to a 4.6% incidence in patients randomized to placebo. 18
1
An assessment by the Statin Intolerance Panel: 2014 update
John R. Guyton, MD, FNLA correspondence email, Harold E. Bays, MD, FNLA, Scott M. Grundy, MD, PhD, FNLA,
Terry A. Jacobson, MD, FACP, FNLA Received: March 3, 2014; Accepted: March 4, 2014;
©Institute for Clinical and Economic Review, 2015
Page 67
Also of note, the precise measurement of statin intolerance is difficult. As noted earlier in our
report, often statin intolerance is primarily associated with muscle pain, which can arise from a
number of other causes, particularly in older individuals. This was echoed by one clinical expert
participant of the Roundtable who conveyed that statin intolerance is primarily a clinical diagnosis;
there is no genetic marker to identify it.
4. Management of patients with possible statin intolerance and other complexities of decisionmaking regarding PCSK9 inhibitors suggest that it is reasonable to restrict prescribing of PCSK9
inhibitors to specialists in lipid management.
Currently, there is variation in how payers define statin intolerance; it will likely take work to define
the concept. One way to ensure that possible statin intolerance and other complexities are
appropriately managed would be to limit prescribing authority of PCSK9 inhibitors to specialists in
lipid management. This was emphasized during the Roundtable when one participant noted that it
is difficult to account for grey areas when releasing a very expensive drug; even among specialists
there is great variability in expertise. Given that cardiologists are likely to see the majority of the
high risk patients it may be reasonable to restrict prescribing of PCSK9 inhibitors to specialists.
5. If the pricing for PCSK9 inhibitors were to fall 50%-85% to a level that aligns with the benefit to
patients and with a reasonable short term affordability, payers would likely consider lifting many
elements of proposed prior authorization requirements.
During the Policy Roundtable, there was a focused discussion on the price of PCSK9 inhibitors as it
relates to other public health expenditures. One participant vocalized that other services are being
starved to pay for these and other drugs. The participant included pointed statements towards the
morality associated with prices charged for new treatments and society’s willingness to pay. There
was an urge for stakeholders to understand how to apply balance between the cost of care versus
other services that must be financed such as public health, education, safety, and other public
programs. When the pharmacy benefit management representative was asked about the price he
responded that if PCSK9 inhibitors were priced lower, the conversation about cost and value would
not be taking place.
The high cost of the pharmaceuticals was also noted by participants both on the payer and provider
side when discussing the administrative burden associated with prior authorization and other
utilization management techniques. These mechanisms are necessary to reduce the budgetary
impact of PCSK9 inhibitors at the current price. However, Roundtable participants noted that if the
drugs were priced near ICER’s value-based price benchmark, the focus on utilization management,
and perhaps the entire discussion surrounding the value and budgetary impact of PCSK9 inhibitors,
would not be taking place. In the future, pharmaceutical companies should join a broader
discussion with payers and other stakeholders about how the current approach to the pricing of
drugs can change to help address the negative impact of rising health care costs on patients and on
other important societal goals.
©Institute for Clinical and Economic Review, 2015
Page 68
6. Future research needs will be strongly influenced by the results of current clinical outcomes
studies, but additional research in adherence and long-term safety will be important to guide
practice and policy in the future.
Discussion during the October 27, 2015 meeting highlighted the need for more information about
efficacy and safety of PCSK9 inhibitors. It was evident that there is an unmet need among patients
with FH, but evidence pertaining to other sub-populations is less clear. When voting on the
evidence, the majority of the council concluded that there was not adequate evidence to
demonstrate that PCSK9 inhibitors improve net health benefits in patients with a history of
cardiovascular disease. In addition to general uncertainty around long-term effectiveness and
safety, most Council members were unable to completely accept the so-called “LDL hypothesis”.
This hypothesis assumes that any lowering of LDL cholesterol would also lower the risk of heart
attacks and strokes. Council members were reluctant to embrace the hypothesis likely because this
has been demonstrated for some drugs (statins, ezetimibe) but not for others (niacin, fibrates). It
will be necessary to revisit the discussion surrounding the direct effects of PCSK9 inhibitors on rates
of heart attack and stroke when results of ongoing trials are released in 2017.
There may also be a need to examine real world data related to difference in treatment adherence
between PCSK9 inhibitors and statins. As noted earlier, it is difficult to measure treatment
adherence among patients who have been prescribed statins. It remains unclear whether treatment
adherence would improve if a patient were to switch from an oral statin to a less frequent, but
injectable, PCSK9 inhibitor. PCSK9 inhibitor adherence was referenced during the evidence review
in the context of clinical trial data. Some felt that there may be differences between adherence
rates observed in RCTs versus those experienced by patients that are not participating in trials.
Some argue rates of adherence found outside of clinical trials will most certainly be lower than
those reported during clinical trials. Adherence rates for PCSK9 inhibitors versus statins will likely
need to be studied further.
There appeared to be consensus among participants of the Roundtable that clinicians should try
multiple statin options before seeking alternative treatments for patients with elevated LDL-C.
There may be opportunity in this area to develop tools to help physicians educate patients about
statin intolerance. One physician shared that patients are often afraid that any experienced muscle
pain while taking statins will be permanent and therefore discontinue use of the medication. As
direct to consumer marketing campaigns are executed for the new LDL-C lowering therapies, statin
intolerance-related muscle pain will likely be used as means of attracting patients to PCSK9
inhibitors. In this environment, it is imperative that patients be equipped with accurate information
pertaining to evidence based therapies in order to make well informed treatment decisions.
©Institute for Clinical and Economic Review, 2015
Page 69
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Navarese EP, Kolodziejczak M, Schulze V, et al. Effects of Proprotein Convertase Subtilisin/Kexin
Type 9 Antibodies in Adults With Hypercholesterolemia: A Systematic Review and Meta-analysis.
Ann Intern Med. 2015.
Zhang XL, Zhu QQ, Zhu L, et al. Safety and efficacy of anti-PCSK9 antibodies: a meta-analysis of
25 randomized, controlled trials. BMC Med. 2015;13:123.
Mozaffarian D, Benjamin EJ, Go AS, et al. Heart disease and stroke statistics--2015 update: a
report from the American Heart Association. Circulation. 2015;131(4):e29-322.
Stone NJ, Robinson JG, Lichtenstein AH, et al. 2013 ACC/AHA guideline on the treatment of
blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: a report of the
American College of Cardiology/American Heart Association Task Force on Practice Guidelines.
Circulation. 2014;129(25 Suppl 2):S1-45.
Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP)
Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults (Adult
Treatment Panel III). Jama. 2001;285(19):2486-2497.
European Association for Cardiovascular P, Rehabilitation, Reiner Z, et al. ESC/EAS Guidelines for
the management of dyslipidaemias: the Task Force for the management of dyslipidaemias of the
European Society of Cardiology (ESC) and the European Atherosclerosis Society (EAS). Eur Heart
J. 2011;32(14):1769-1818.
Shrank WH, Barlow JF, Brennan TA. New Therapies in the Treatment of High Cholesterol: An
Argument to Return to Goal-Based Lipid Guidelines. JAMA; 2015.
Cho L, Rocco M, Colquhoun D, et al. Design and rationale of the GAUSS-2 study trial: a doubleblind, ezetimibe-controlled phase 3 study of the efficacy and tolerability of evolocumab (AMG
145) in subjects with hypercholesterolemia who are intolerant of statin therapy. Clinical
cardiology. 2014;37(3):131-139.
Liyanage KE, Burnett JR, Hooper AJ, van Bockxmeer FM. Familial hypercholesterolemia:
epidemiology, Neolithic origins and modern geographic distribution. Critical reviews in clinical
laboratory sciences. 2011;48(1):1-18.
Page MM, Stefanutti C, Sniderman A, Watts GF. Recent advances in the understanding and care
of familial hypercholesterolaemia: significance of the biology and therapeutic regulation of
proprotein convertase subtilisin/kexin type 9. Clinical science (London, England : 1979).
2015;129(1):63-79.
Sniderman AD, Tsimikas S, Fazio S. The severe hypercholesterolemia phenotype: clinical
diagnosis, management, and emerging therapies. Journal of the American College of Cardiology.
2014;63(19):1935-1947.
Vishwanath R, Hemphill LC. Familial hypercholesterolemia and estimation of US patients eligible
for low-density lipoprotein apheresis after maximally tolerated lipid-lowering therapy. Journal of
clinical lipidology. 2014;8(1):18-28.
Sjouke B, Kusters DM, Kindt I, et al. Homozygous autosomal dominant hypercholesterolaemia in
the Netherlands: prevalence, genotype-phenotype relationship, and clinical outcome. Eur Heart
J. 2015;36(9):560-565.
de Ferranti S, Rodday A, Mendelson M, Wong J, Leslie L, Sheldrick R. Abstract 19656: What is the
Prevalence of Familial Hypercholesterolemia in the US? Vol 130: Circulation; 2014:A19656.
©Institute for Clinical and Economic Review, 2015
Page 70
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
Banach M, Rizzo M, Toth PP, et al. Statin intolerance - an attempt at a unified definition. Position
paper from an International Lipid Expert Panel. Archives of medical science : AMS. 2015;11(1):123.
Stroes ES, Thompson PD, Corsini A, et al. Statin-associated muscle symptoms: impact on statin
therapy-European Atherosclerosis Society Consensus Panel Statement on Assessment, Aetiology
and Management. Eur Heart J. 2015;36(17):1012-1022.
Bruckert E, Hayem G, Dejager S, Yau C, Begaud B. Mild to moderate muscular symptoms with
high-dosage statin therapy in hyperlipidemic patients--the PRIMO study. Cardiovascular drugs
and therapy / sponsored by the International Society of Cardiovascular Pharmacotherapy.
2005;19(6):403-414.
Parker BA, Capizzi JA, Grimaldi AS, et al. Effect of statins on skeletal muscle function. Circulation.
2013;127(1):96-103.
Mikhailidis DP, Sibbring GC, Ballantyne CM, Davies GM, Catapano AL. Meta-analysis of the
cholesterol-lowering effect of ezetimibe added to ongoing statin therapy. Current medical
research and opinion. 2007;23(8):2009-2026.
Pandor A, Ara RM, Tumur I, et al. Ezetimibe monotherapy for cholesterol lowering in 2,722
people: systematic review and meta-analysis of randomized controlled trials. J Intern Med.
2009;265(5):568-580.
Kastelein JJ, Akdim F, Stroes ES, et al. Simvastatin with or without ezetimibe in familial
hypercholesterolemia. The New England journal of medicine. 2008;358(14):1431-1443.
Villines TC, Stanek EJ, Devine PJ, et al. The ARBITER 6-HALTS Trial (Arterial Biology for the
Investigation of the Treatment Effects of Reducing Cholesterol 6-HDL and LDL Treatment
Strategies in Atherosclerosis): final results and the impact of medication adherence, dose, and
treatment duration. Journal of the American College of Cardiology. 2010;55(24):2721-2726.
Bays H, Farnier M, Gaudet D, et al. Efficacy and Safety of Combining Alirocumab With
Atorvastatin or Rosuvastatin versus adding ezetimibe, doubling statin dose or switching statin
therapy in high cardiovascular risk patients: ODYSSEY OPTIONS I and II. Paper presented at:
American Heart Association Scientific Sessions; November 15, 2014, 2014; Chicago, Illinois.
Bays H, Gaudet D, Weiss R, et al. Alirocumab as Add-On to Atorvastatin Versus Other Lipid
Treatment Strategies: ODYSSEY OPTIONS I Randomized Trial. J Clin Endocrinol Metab.
2015;100(8):3140-3148.
Blom DJ, Hala T, Bolognese M, et al. A 52-week placebo-controlled trial of evolocumab in
hyperlipidemia. The New England journal of medicine. 2014;370(19):1809-1819.
Cannon CP, Cariou B, Blom D, et al. Efficacy and safety of alirocumab in high cardiovascular risk
patients with inadequately controlled hypercholesterolaemia on maximally tolerated doses of
statins: the ODYSSEY COMBO II randomized controlled trial. Eur Heart J. 2015.
Colhoun HM, Robinson JG, Farnier M, et al. Efficacy and safety of alirocumab, a fully human
PCSK9 monoclonal antibody, in high cardiovascular risk patients with poorly controlled
hypercholesterolemia on maximally tolerated doses of statins: rationale and design of the
ODYSSEY COMBO I and II trials. BMC Cardiovasc Disord. 2014;14:121.
Desai NR, Giugliano RP, Zhou J, et al. AMG 145, a monoclonal antibody against PCSK9, facilitates
achievement of national cholesterol education program-adult treatment panel III low-density
lipoprotein cholesterol goals among high-risk patients: an analysis from the LAPLACE-TIMI 57
trial (LDL-C assessment with PCSK9 monoclonal antibody inhibition combined with statin
thErapy-thrombolysis in myocardial infarction 57). Journal of the American College of
Cardiology. 2014;63(5):430-433.
Desai NR, Kohli P, Giugliano RP, et al. AMG145, a monoclonal antibody against proprotein
convertase subtilisin kexin type 9, significantly reduces lipoprotein(a) in hypercholesterolemic
©Institute for Clinical and Economic Review, 2015
Page 71
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
patients receiving statin therapy: an analysis from the LDL-C Assessment with Proprotein
Convertase Subtilisin Kexin Type 9 Monoclonal Antibody Inhibition Combined with Statin
Therapy (LAPLACE)-Thrombolysis in Myocardial Infarction (TIMI) 57 trial. Circulation.
2013;128(9):962-969.
Gaudet D, Kereiakes DJ, McKenney JM, et al. Effect of alirocumab, a monoclonal proprotein
convertase subtilisin/kexin 9 antibody, on lipoprotein(a) concentrations (a pooled analysis of
150 mg every two weeks dosing from phase 2 trials). Am J Cardiol. 2014;114(5):711-715.
Ginsberg HN, Rader D, Raal F, Guyton JR, Lorenzato C, Pordy R. ODYSSEY HIGH FH: Efficacy and
safety of alirocumab in patients with severe heterozygous familial hypercholesterolemia. Paper
presented at: American Heart Association Scientific Sessions; November 15, 2014, 2014;
Chicago, Illinois.
Giugliano RP, Desai NR, Kohli P, et al. Efficacy, safety, and tolerability of a monoclonal antibody
to proprotein convertase subtilisin/kexin type 9 in combination with a statin in patients with
hypercholesterolaemia (LAPLACE-TIMI 57): a randomised, placebo-controlled, dose-ranging,
phase 2 study. Lancet. 2012;380(9858):2007-2017.
Hirayama A, Honarpour N, Yoshida M, et al. Effects of evolocumab (AMG 145), a monoclonal
antibody to PCSK9, in hypercholesterolemic, statin-treated Japanese patients at high
cardiovascular risk--primary results from the phase 2 YUKAWA study. Circ J. 2014;78(5):10731082.
Kastelein JJ, Ginsberg HN, Langslet G, et al. Efficacy and safety of alircoumab in patients with
heterozygous familial hypercholesterolemia (heFH) not adequately controlled with current lipidlowering therapy: results of ODYSSEY FH I and FH II studies. European Society of Cardiology
Congress; August 30, 2014, 2014; Barcelona, Spain.
Kastelein JJ, Robinson JG, Farnier M, et al. Efficacy and safety of alirocumab in patients with
heterozygous familial hypercholesterolemia not adequately controlled with current lipidlowering therapy: design and rationale of the ODYSSEY FH studies. Cardiovascular drugs and
therapy / sponsored by the International Society of Cardiovascular Pharmacotherapy.
2014;28(3):281-289.
Kereiakes DJ, Robinson JG, Cannon CP, et al. Efficacy and safety of the proprotein convertase
subtilisin/kexin type 9 inhibitor alirocumab among high cardiovascular risk patients on
maximally tolerated statin therapy: The ODYSSEY COMBO I study. Am Heart J. 2015;169(6):906915 e913.
Kohli P, Desai NR, Giugliano RP, et al. Design and rationale of the LAPLACE-TIMI 57 trial: a phase
II, double-blind, placebo-controlled study of the efficacy and tolerability of a monoclonal
antibody inhibitor of PCSK9 in subjects with hypercholesterolemia on background statin therapy.
Clinical cardiology. 2012;35(7):385-391.
Koren MJ, Giugliano RP, Raal FJ, et al. Efficacy and safety of longer-term administration of
evolocumab (AMG 145) in patients with hypercholesterolemia: 52-week results from the OpenLabel Study of Long-Term Evaluation Against LDL-C (OSLER) randomized trial. Circulation.
2014;129(2):234-243.
Koren MJ, Lundqvist P, Bolognese M, et al. Anti-PCSK9 monotherapy for hypercholesterolemia:
the MENDEL-2 randomized, controlled phase III clinical trial of evolocumab. Journal of the
American College of Cardiology. 2014;63(23):2531-2540.
Koren MJ, Roth EM, McKenney JM, et al. Safety and efficacy of alirocumab 150 mg every 2
weeks, a fully human proprotein convertase subtilisin/kexin type 9 monoclonal antibody: A
Phase II pooled analysis. Postgrad Med. 2015;127(2):125-132.
Koren MJ, Scott R, Kim JB, et al. Efficacy, safety, and tolerability of a monoclonal antibody to
proprotein convertase subtilisin/kexin type 9 as monotherapy in patients with
©Institute for Clinical and Economic Review, 2015
Page 72
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
hypercholesterolaemia (MENDEL): a randomised, double-blind, placebo-controlled, phase 2
study. Lancet. 2012;380(9858):1995-2006.
McKenney JM, Koren MJ, Kereiakes DJ, Hanotin C, Ferrand AC, Stein EA. Safety and efficacy of a
monoclonal antibody to proprotein convertase subtilisin/kexin type 9 serine protease,
SAR236553/REGN727, in patients with primary hypercholesterolemia receiving ongoing stable
atorvastatin therapy. Journal of the American College of Cardiology. 2012;59(25):2344-2353.
Moriarty PM, Jacobson TA, Bruckert E, et al. Efficacy and safety of alirocumab, a monoclonal
antibody to PCSK9, in statin-intolerant patients: design and rationale of ODYSSEY ALTERNATIVE,
a randomized phase 3 trial. Journal of clinical lipidology. 2014;8(6):554-561.
Moriarty PM, Thompson PD, Cannon CP, et al. ODYSSEY ALTERNATIVE: Efficacy and safety of
alirocumab versus ezetimibe in patients with statin intlerance defined by placebo run-in and
statin rechallenge arm. American Heart Association Scientific Sessions; November 15, 2014,
2014; Chicago, Illinois.
Raal F, Scott R, Somaratne R, et al. Low-density lipoprotein cholesterol-lowering effects of AMG
145, a monoclonal antibody to proprotein convertase subtilisin/kexin type 9 serine protease in
patients with heterozygous familial hypercholesterolemia: the Reduction of LDL-C with PCSK9
Inhibition in Heterozygous Familial Hypercholesterolemia Disorder (RUTHERFORD) randomized
trial. Circulation. 2012;126(20):2408-2417.
Raal FJ, Giugliano RP, Sabatine MS, et al. Reduction in lipoprotein(a) with PCSK9 monoclonal
antibody evolocumab (AMG 145): a pooled analysis of more than 1,300 patients in 4 phase II
trials. Journal of the American College of Cardiology. 2014;63(13):1278-1288.
Raal FJ, Honarpour N, Blom DJ, et al. Inhibition of PCSK9 with evolocumab in homozygous
familial hypercholesterolaemia (TESLA Part B): a randomised, double-blind, placebo-controlled
trial. Lancet. 2015;385(9965):341-350.
Raal FJ, Stein EA, Dufour R, et al. PCSK9 inhibition with evolocumab (AMG 145) in heterozygous
familial hypercholesterolaemia (RUTHERFORD-2): a randomised, double-blind, placebocontrolled trial. Lancet. 2015;385(9965):331-340.
Robinson JG, Colhoun HM, Bays HE, et al. Efficacy and safety of alirocumab as add-on therapy in
high-cardiovascular-risk patients with hypercholesterolemia not adequately controlled with
atorvastatin (20 or 40 mg) or rosuvastatin (10 or 20 mg): design and rationale of the ODYSSEY
OPTIONS Studies. Clinical cardiology. 2014;37(10):597-604.
Robinson JG, Farnier M, Krempf M, et al. Efficacy and safety of alirocumab in reducing lipids and
cardiovascular events. The New England journal of medicine. 2015;372(16):1489-1499.
Robinson JG, Nedergaard BS, Rogers WJ, et al. Effect of evolocumab or ezetimibe added to
moderate- or high-intensity statin therapy on LDL-C lowering in patients with
hypercholesterolemia: the LAPLACE-2 randomized clinical trial. Jama. 2014;311(18):1870-1882.
Robinson JG, Rogers WJ, Nedergaard BS, et al. Rationale and design of LAPLACE-2: a phase 3,
randomized, double-blind, placebo- and ezetimibe-controlled trial evaluating the efficacy and
safety of evolocumab in subjects with hypercholesterolemia on background statin therapy.
Clinical cardiology. 2014;37(4):195-203.
Roth EM, McKenney JM. ODYSSEY MONO: effect of alirocumab 75 mg subcutaneously every 2
weeks as monotherapy versus ezetimibe over 24 weeks. Future Cardiol. 2015;11(1):27-37.
Roth EM, McKenney JM, Hanotin C, Asset G, Stein EA. Atorvastatin with or without an antibody
to PCSK9 in primary hypercholesterolemia. The New England journal of medicine.
2012;367(20):1891-1900.
Roth EM, Taskinen MR, Ginsberg HN, et al. Monotherapy with the PCSK9 inhibitor alirocumab
versus ezetimibe in patients with hypercholesterolemia: results of a 24 week, double-blind,
randomized Phase 3 trial. Int J Cardiol. 2014;176(1):55-61.
©Institute for Clinical and Economic Review, 2015
Page 73
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
Sabatine MS, Giugliano RP, Wiviott SD, et al. Efficacy and safety of evolocumab in reducing lipids
and cardiovascular events. The New England journal of medicine. 2015;372(16):1500-1509.
Schwartz GG, Bessac L, Berdan LG, et al. Effect of alirocumab, a monoclonal antibody to PCSK9,
on long-term cardiovascular outcomes following acute coronary syndromes: rationale and
design of the ODYSSEY outcomes trial. Am Heart J. 2014;168(5):682-689.
Stein EA, Gipe D, Bergeron J, et al. Effect of a monoclonal antibody to PCSK9,
REGN727/SAR236553, to reduce low-density lipoprotein cholesterol in patients with
heterozygous familial hypercholesterolaemia on stable statin dose with or without ezetimibe
therapy: a phase 2 randomised controlled trial. Lancet. 2012;380(9836):29-36.
Stein EA, Giugliano RP, Koren MJ, et al. Efficacy and safety of evolocumab (AMG 145), a fully
human monoclonal antibody to PCSK9, in hyperlipidaemic patients on various background lipid
therapies: pooled analysis of 1359 patients in four phase 2 trials. Eur Heart J. 2014;35(33):22492259.
Stein EA, Mellis S, Yancopoulos GD, et al. Effect of a monoclonal antibody to PCSK9 on LDL
cholesterol. The New England journal of medicine. 2012;366(12):1108-1118.
Stroes E, Colquhoun D, Sullivan D, et al. Anti-PCSK9 antibody effectively lowers cholesterol in
patients with statin intolerance: the GAUSS-2 randomized, placebo-controlled phase 3 clinical
trial of evolocumab. Journal of the American College of Cardiology. 2014;63(23):2541-2548.
Sullivan D, Olsson AG, Scott R, et al. Effect of a monoclonal antibody to PCSK9 on low-density
lipoprotein cholesterol levels in statin-intolerant patients: the GAUSS randomized trial. Jama.
2012;308(23):2497-2506.
Bibbins-Domingo K, Coxson P, Pletcher MJ, Lightwood J, Goldman L. Adolescent overweight and
future adult coronary heart disease. The New England journal of medicine. 2007;357(23):23712379.
Hunink MGM, Goldman L, Tosteson ANA, et al. The recent decline in mortality from coronary
heart disease, 1980-1990: the effect of secular trends in risk factors and treatment. Jama.
1997;277(7):535-542.
Weinstein MC, Coxson PG, Williams LW, Pass TM, Stason WB, Goldman L. Forecasting coronary
heart disease incidence, mortality, and cost: the Coronary Heart Disease Policy Model. Am J of
Public Health. 1987;77(11):1417-1426.
Odden MC, Pletcher MJ, Coxson PG, et al. Cost-effectiveness and population impact of statins
for primary prevention in adults aged 75 years or older in the United States. Ann Intern Med.
2015;162(8):533-541.
Moran AE, Odden MC, Thanataveerat A, et al. Cost-effectiveness of hypertension therapy
according to 2014 guidelines. The New England journal of medicine. 2015;372(5):447-455.
Konfino J, Ferrante D, Mejia R, et al. Impact on cardiovascular disease events of the
implementation of Argentina's national tobacco control law. Tob Control. 2014;23(2):e6.
Wang YC, Coxson P, Shen YM, Goldman L, Bibbins-Domingo K. A penny-per-ounce tax on sugarsweetened beverages would cut health and cost burdens of diabetes. Health Aff (Millwood).
2012;31(1):199-207.
Ferrante D, Konfino J, Mejia R, et al. [The cost-utility ratio of reducing salt intake and its impact
on the incidence of cardiovascular disease in Argentina]. Rev Panam Salud Publica.
2012;32(4):274-280.
Lazar LD, Pletcher MJ, Coxson PG, Bibbins-Domingo K, Goldman L. Cost-effectiveness of statin
therapy for primary prevention in a low-cost statin era. Circulation. 2011;124(2):146-153.
Bibbins-Domingo K, Chertow GM, Coxson PG, et al. Projected effect of dietary salt reductions on
future cardiovascular disease. The New England journal of medicine. 2010;362(7):590-599.
©Institute for Clinical and Economic Review, 2015
Page 74
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
Pletcher MJ, Lazar L, Bibbins-Domingo K, et al. Comparing impact and cost-effectiveness of
primary prevention strategies for lipid-lowering. Ann Intern Med. 2009;150(4):243-254.
Gaspoz JM, Coxson PG, Goldman PA, et al. Cost effectiveness of aspirin, clopidogrel, or both for
secondary prevention of coronary heart disease. The New England journal of medicine.
2002;346(23):1800-1806.
Phillips KA, Shlipak MG, Coxson P, et al. Health and economic benefits of increased beta-blocker
use following myocardial infarction. Jama. 2000;284(21):2748-2754.
Torrance GW, Siegel JE, Luce BR. Framing and designing the cost-effectiveness analysis.
In: Gold MR, Siegel JE, Rusell LB, eds. Cost-Effectiveness in Health and Medicine. . New
York, NY: Oxford University Press; 1996:54-81.
U.S. National Library of Medicine. Label: Praluent - alirocumab injection, solution. 2015;
http://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=446f6b5c-0dd4-44ff-9bc2c2b41f2806b4. Accessed 9/2/2015, 2015.
Cannon CP, Blazing MA, Giugliano RP, et al. Ezetimibe Added to Statin Therapy after Acute
Coronary Syndromes. N Engl J Med. 2015;372(25):2387-2397.
Anderson JL, Heidenreich PA, Barnett PG, et al. ACC/AHA statement on cost/value methodology
in clinical practice guidelines and performance measures: a report of the American College of
Cardiology/American Heart Association Task Force on Performance Measures and Task Force on
Practice Guidelines. Journal of the American College of Cardiology. 2014;63(21):2304-2322.
Goldstein JL, Brown MS. A century of cholesterol and coronaries: from plaques to genes to
statins. Cell. 2015;161(1):161-172.
Baigent C, Blackwell L, Emberson J, et al. Efficacy and safety of more intensive lowering of LDL
cholesterol: a meta-analysis of data from 170,000 participants in 26 randomised trials. Lancet.
2010;376(9753):1670-1681.
Fulcher J, O'Connell R, Voysey M, et al. Efficacy and safety of LDL-lowering therapy among men
and women: meta-analysis of individual data from 174,000 participants in 27 randomised trials.
Lancet. 2015;385(9976):1397-1405.
Taylor F, Huffman MD, Macedo AF, et al. Statins for the primary prevention of cardiovascular
disease. Cochrane Database Syst Rev. 2013;1:CD004816.
Ben-Yehuda O, DeMaria AN. LDL-cholesterol targets after the ACC/AHA 2013 guidelines:
evidence that lower is better? Journal of the American College of Cardiology. 2014;64(5):495497.
Filippatos TD, Elisaf MS. Are lower levels of LDL-cholesterol really better? Looking at the results
of IMPROVE-IT: opinions of three experts - III. Hellenic journal of cardiology : HJC = Hellenike
kardiologike epitheorese. 2015;56(1):7-9.
Jarcho JA, Keaney JF, Jr. Proof That Lower Is Better--LDL Cholesterol and IMPROVE-IT. The New
England journal of medicine. 2015;372(25):2448-2450.
Jones PH. Lipid-lowering treatment in coronary artery disease: how low should cholesterol go?
Drugs. 2000;59(5):1127-1135.
Stein EA, Raal FJ. Targeting LDL: is lower better and is it safe? Best practice & research. Clinical
endocrinology & metabolism. 2014;28(3):309-324.
The Coronary Drug Project. Findings leading to discontinuation of the 2.5-mg day estrogen
group. The coronary Drug Project Research Group. Jama. 1973;226(6):652-657.
Barter PJ, Caulfield M, Eriksson M, et al. Effects of torcetrapib in patients at high risk for
coronary events. The New England journal of medicine. 2007;357(21):2109-2122.
©Institute for Clinical and Economic Review, 2015
Page 75
91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
107.
Boden WE, Probstfield JL, Anderson T, et al. Niacin in patients with low HDL cholesterol levels
receiving intensive statin therapy. The New England journal of medicine. 2011;365(24):22552267.
Hulley S, Grady D, Bush T, et al. Randomized trial of estrogen plus progestin for secondary
prevention of coronary heart disease in postmenopausal women. Heart and Estrogen/progestin
Replacement Study (HERS) Research Group. Jama. 1998;280(7):605-613.
Manson JE, Hsia J, Johnson KC, et al. Estrogen plus progestin and the risk of coronary heart
disease. The New England journal of medicine. 2003;349(6):523-534.
Dadu RT, Ballantyne CM. Lipid lowering with PCSK9 inhibitors. Nature reviews. Cardiology.
2014;11(10):563-575.
Khayznikov M, Hemachrandra K, Pandit R, Kumar A, Wang P, Glueck CJ. Statin Intolerance
Because of Myalgia, Myositis, Myopathy, or Myonecrosis Can in Most Cases be Safely Resolved
by Vitamin D Supplementation. North American journal of medical sciences. 2015;7(3):86-93.
Abifadel M, Rabes JP, Devillers M, et al. Mutations and polymorphisms in the proprotein
convertase subtilisin kexin 9 (PCSK9) gene in cholesterol metabolism and disease. Human
mutation. 2009;30(4):520-529.
Abifadel M, Bernier L, Dubuc G, et al. A PCSK9 variant and familial combined hyperlipidaemia.
Journal of medical genetics. 2008;45(12):780-786.
Abifadel M, Guerin M, Benjannet S, et al. Identification and characterization of new gain-offunction mutations in the PCSK9 gene responsible for autosomal dominant
hypercholesterolemia. Atherosclerosis. 2012;223(2):394-400.
Abifadel M, Varret M, Rabes JP, et al. Mutations in PCSK9 cause autosomal dominant
hypercholesterolemia. Nature genetics. 2003;34(2):154-156.
Humphries SE, Cranston T, Allen M, et al. Mutational analysis in UK patients with a clinical
diagnosis of familial hypercholesterolaemia: relationship with plasma lipid traits, heart disease
risk and utility in relative tracing. Journal of molecular medicine (Berlin, Germany).
2006;84(3):203-214.
Humphries SE, Whittall RA, Hubbart CS, et al. Genetic causes of familial hypercholesterolaemia
in patients in the UK: relation to plasma lipid levels and coronary heart disease risk. Journal of
medical genetics. 2006;43(12):943-949.
Cariou B, Ouguerram K, Zair Y, et al. PCSK9 dominant negative mutant results in increased LDL
catabolic rate and familial hypobetalipoproteinemia. Arteriosclerosis, thrombosis, and vascular
biology. 2009;29(12):2191-2197.
Cohen JC, Boerwinkle E, Mosley TH, Jr., Hobbs HH. Sequence variations in PCSK9, low LDL, and
protection against coronary heart disease. The New England journal of medicine.
2006;354(12):1264-1272.
Fasano T, Cefalu AB, Di Leo E, et al. A novel loss of function mutation of PCSK9 gene in white
subjects with low-plasma low-density lipoprotein cholesterol. Arteriosclerosis, thrombosis, and
vascular biology. 2007;27(3):677-681.
Benn M, Nordestgaard BG, Grande P, Schnohr P, Tybjaerg-Hansen A. PCSK9 R46L, low-density
lipoprotein cholesterol levels, and risk of ischemic heart disease: 3 independent studies and
meta-analyses. Journal of the American College of Cardiology. 2010;55(25):2833-2842.
Methods Guide for Effectiveness and Comparative Effectiveness Reviews. 2014.
http://effectivehealthcare.ahrq.gov/index.cfm/search-for-guides-reviews-andreports/?productid=318&pageaction=displayproduct. Accessed March 2015.
Lipscomb J, Weinstein MC, Torrance GW. Time preference. In: Gold MR, Siegel JE, Russel LB, eds.
Cost-Effectiveness in Health and Medicine. New York: Oxford University Press; 1996:214-246.
©Institute for Clinical and Economic Review, 2015
Page 76
108.
109.
110.
111.
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
125.
126.
127.
Weinstein MC, Siegel JE, Gold MR, Kamlet MS, Russell LB. Recommendations of the Panel on
Cost-effectiveness in Health and Medicine. Jama. 1996;276(15):1253-1258.
Cholesterol Treatment Trialists C, Baigent C, Blackwell L, et al. Efficacy and safety of more
intensive lowering of LDL cholesterol: a meta-analysis of data from 170,000 participants in 26
randomised trials. Lancet. 2010;376(9753):1670-1681.
Shepherd J, Cobbe SM, Ford I, et al. Prevention of coronary heart disease with pravastatin in
men with hypercholesterolemia. West of Scotland Coronary Prevention Study Group. The New
England journal of medicine. 1995;333(20):1301-1307.
Goldberg AC, Hopkins PN, Toth PP, et al. Familial hypercholesterolemia: screening, diagnosis and
management of pediatric and adult patients: clinical guidance from the National Lipid
Association Expert Panel on Familial Hypercholesterolemia. Journal of clinical lipidology.
2011;5(3 Suppl):S1-8.
Singh S, Bittner V. Familial hypercholesterolemia--epidemiology, diagnosis, and screening. Curr
Atheroscler Rep. 2015;17(2):482.
Fernandez G, Spatz ES, Jablecki C, Phillips PS. Statin myopathy: a common dilemma not reflected
in clinical trials. Cleve Clin J Med. 2011;78(6):393-403.
Grundy SM. Statin discontinuation and intolerance: the challenge of lifelong therapy. Ann Intern
Med. 2013;158(7):562-563.
Zhang H, Plutzky J, Skentzos S, et al. Discontinuation of statins in routine care settings: a cohort
study. Ann Intern Med. 2013;158(7):526-534.
Ara R, Tumur I, Pandor A, et al. Ezetimibe for the treatment of hypercholesterolaemia: a
systematic review and economic evaluation. Health Technol Assess. 2008;12(21):iii, xi-xiii, 1-212.
Randers MB, Andersen JL, Petersen J, et al. Exercise performance and cardiovascular health
variables in 70-year-old male soccer players compared to endurance-trained, strength-trained
and untrained age-matched men. J Sports Sci. 2014;32(13):1300-1308.
California Office of Statewide Health Planning & Development. Hospital financial data for cost to
charge ratio, CA inpatinet discharage data hospital annual financial data, pivot profiles. 19992000; http://www.oshpd.ca.gov/HQAD/Hospital/financial/hospAF.htm Accessed June, 2003.
Average cost to community hospitals per patient, by state (Table 204). Statistical Abstract of the
United States. Bureau of the Census. Washington, D.C.: Government Printing Office; 1998:136.
Medical Expenditure Panel Survey. Medical Expenditure Panel Survey Public Use Files 19982008. http://meps.ahrq.gov/mepsweb/, 2010.
Moran AE, Forouzanfar MH, Roth G, et al. Temporal Trends in Ischemic Heart Disease Mortality
in 21 World Regions, 1980-2010: The Global Burden of Disease 2010 Study. Circulation. 2014.
Moran AE, Forouzanfar MH, Roth G, et al. The Global Burden of Ischemic Heart Disease in 1990
and 2010: The Global Burden of Disease 2010 Study. Circulation. 2014.
Murray CJ, Vos T, Lozano R, et al. Disability-adjusted life years (DALYs) for 291 diseases and
injuries in 21 regions, 1990-2010: a systematic analysis for the Global Burden of Disease Study
2010. Lancet. 2012;380(9859):2197-2223.
Amarenco P, Bogousslavsky J, Callahan A, 3rd, et al. High-dose atorvastatin after stroke or
transient ischemic attack. The New England journal of medicine. 2006;355(6):549-559.
Avorn J. The $2.6 billion pill--methodologic and policy considerations. The New England journal
of medicine. 2015;372(20):1877-1879.
Siddiqui M, Rajkumar SV. The high cost of cancer drugs and what we can do about it. Mayo Clin
Proc. 2012;87(10):935-943.
Reinhardt U. JAMA Forum: Probing our Moral Values in Health Care: the Pricing of Specialty
Drugs. Accessed at http://newsatjama.jama.com/2015/08/11/jama-forum-probing-our-moralvalues-in-health-care-the-pricing-of-specialty-drugs/ on August 17, 2015. 2015.
©Institute for Clinical and Economic Review, 2015
Page 77
128.
129.
130.
131.
132.
133.
134.
135.
136.
137.
138.
139.
140.
141.
142.
143.
144.
Osterberg L, Blaschke T. Adherence to medication. The New England journal of medicine.
2005;353(5):487-497.
U.S. Census Bureau Population Division. Monthly Postcensal Resident Population by Single Year
of Age, Sex, Race and Hispanic Origin for the United States: July 1, 2010 to December 1 2010
(NC-EST2011-ALLDATA-R-File02). 2011;
http://www.census.gov/popest/data/national/asrh/2011/files/NC-EST2011-ALLDATA-RFile02.csv. Accessed January 15, 2013.
U.S. Census Bureau Population Division. Projected Population by Single Year of Age, Sex, Race,
and Hispanic Origin for the United States: 2012 to 2060 (NP2012_D1). 2012;
http://www.census.gov/population/projections/files/downloadables/NP2012_D1.csv Accessed
January 15, 2013.
U.S. Census Bureau Population Division. Methodology and Assumptions for the 2012 National
Projections. 2012; December
2012:http://www.census.gov/population/projections/files/methodology/methodstatement12.p
df. Accessed January 15, 2013.
Centers for Disease Control and Prevention; National Center for Health Statistics. Underlying
cause of death 1999-2010 on CDC WONDER online database, released 2012. 2012;
http://wonder.cdc.gov/ucd-icd10.html. Accessed January 15, 2013.
National Center for Health Statistics. ICD10 Codes. 2004;
http://ftp.cdc.gov/pub/Health_Statistics/NCHS/Publications/ICD10/each10.txt. Accessed July
12, 2006.
Consensus recommendations for the management of chronic heart failure. On behalf of the
membership of the advisory council to improve outcomes nationwide in heart failure. Am J
Cardiol. 1999;83:1A-38A.
Dawber TR. The Framingham Study: the epidemiology of atherosclerotic disease. Cambridge,
MA: Harvard University Press; 1980.
Feinleib M, Kannel WB, Garrison RJ, McNamara PM, Castelli WP. The Framingham Offspring
Study. Design and preliminary data. Preventive medicine. 1975;4(4):518-525.
National Center for Health Statistics. National Health and Nutrition Examination Survey, 20092012. http://wwwn.cdc.gov/nchs/nhanes/search/nhanes09_10.aspx &
http://wwwn.cdc.gov/nchs/nhanes/search/nhanes11_12.aspx, 2012.
Brindle P, Emberson J, Lampe F, et al. Predictive accuracy of the Framingham coronary risk score
in British men: prospective cohort study. BMJ. 2003;327:1267-1270.
D'Agostino RB, Grundy S, Sullivan LM, Wilson PW. Validation of the Framingham Coronary Heart
Disease Prediction Scores. Jama. 2001;286:180-187.
Liu J, Hong Y, D'Agostino RBS, et al. Predictive value for the Chinese population of the
Framingham CHD risk assessment tool compared with the Chinese Multi-Provincial Cohort
Study. Jama. 2004;291(21):2591-2599.
Wilson PW, D'Agostino RB, Levy D, Belanger AM, Silbershaltz H, Kannel WB. Prediction of
coronary heart disease using risk factor categories. Circulation. 1998;97:1837-1847.
Parish S, Collins R, Peto R, et al. Cigarette smoking, tar yields, and non-fatal myocardial
infarction: 14,000 cases and 32,000 controls in the United Kingdom. The International Studies of
Infarct Survival (ISIS) Collaborators. BMJ. 1995;311(7003):471-477.
Law MR, Morris JK, Wald NJ. Environmental tobacco smoke exposure and ischaemic heart
disease: an evaluation of the evidence. BMJ. 1997;315(7114):973-980.
National Center for Health Statistics. National Hospital Discharge Survey.
ftp://ftp.cdc.gov/pub/Health_Statistics/NCHS/Datasets/NHDS/ Accessed March 29, 2012.
©Institute for Clinical and Economic Review, 2015
Page 78
145.
146.
147.
148.
149.
150.
151.
152.
153.
154.
155.
156.
157.
158.
159.
160.
161.
162.
163.
Petersen LA, Wright S, Normand S-LT, Daley J. Positive predictive value of the diagnosis of acute
myocardial infarction in an administrative database. J Gen Intern Med. 1999;14:555-558.
Canto JG, Rogers WJ, Goldberg RJ, et al. Association of age and sex with myocardial infarction
symptom presentation and in-hospital mortality. Jama. 2012;307(8):813-822.
Rieves D, Wright G, Gupta G, Shacter E. Clinical trial (GUSTO-1 and INJECT) evidence of earlier
death for men than women after acute myocardial infarction. Am J Cardiol. 2000;85(2):147-153.
Vaccarino V, Parsons L, Peterson ED, Rogers WJ, Kiefe CI, Canto J. Sex differences in mortality
after acute myocardial infarction: changes from 1994 to 2006. Archives of internal medicine.
2009;169(19):1767-1774.
Medical Expenditure Panel Survey. Medical Expenditure Panel Survey Public Use Files 19962001. http://www.meps.ahrq.gov/Puf/PufSearch.asp?SearchOption=Keyword March 2004 February 2005.
Groeneveld PW, Heidenreich PA, Garber AM. Racial disparity in cardiac procedures and
mortality among long-term survivors of cardiac arrest. Circulation. 2003;108:286-291.
Rea TD, Crouthamel M, Eisenberg MS, Becker LJ, Lima AR. Temporal patterns in long-term
survival after resuscitation from out-of-hospital cardiac arrest. Circulation. 2003;108:1196-1201.
Williams GR, Jiang JG, Matchar DB, Samsa GP. Incidence and occurrence of total (first-ever and
recurrent) stroke. Stroke; a journal of cerebral circulation. 1999;30(12):2523-2528.
The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. The ARIC
investigators. American journal of epidemiology. 1989;129(4):687-702.
Rosamond WD, Folsom AR, Chambless LE, et al. Stroke incidence and survival among middleaged adults: 9-year follow-up of the Atherosclerosis Risk in Communities (ARIC) cohort. Stroke; a
journal of cerebral circulation. 1999;30(4):736-743.
Go AS, Mozaffarian D, Roger VL, et al. Heart disease and stroke statistics--2014 update: a report
from the American Heart Association. Circulation. 2014;129(3):e28-e292.
Witt BJ, Brown RD, Jr., Jacobsen SJ, Weston SA, Yawn BP, Roger VL. A community-based study of
stroke incidence after myocardial infarction. Ann Intern Med. 2005;143(11):785-792.
Appelros P, Gunnarsson KE, Terent A. Ten-year risk for myocardial infarction in patients with
first-ever stroke: a community-based study. Acta neurologica Scandinavica. 2011;124(6):383389.
Behar S, Tanne D, Abinader E, et al. Cerebrovascular accident complicating acute myocardial
infarction: incidence, clinical significance and short- and long-term mortality rates. The SPRINT
Study Group. The American journal of medicine. 1991;91(1):45-50.
Lakshminarayan K, Schissel C, Anderson DC, et al. Five-year rehospitalization outcomes in a
cohort of patients with acute ischemic stroke: Medicare linkage study. Stroke; a journal of
cerebral circulation. 2011;42(6):1556-1562.
Prosser J, MacGregor L, Lees KR, Diener HC, Hacke W, Davis S. Predictors of early cardiac
morbidity and mortality after ischemic stroke. Stroke; a journal of cerebral circulation.
2007;38(8):2295-2302.
Touze E, Varenne O, Chatellier G, Peyrard S, Rothwell PM, Mas JL. Risk of myocardial infarction
and vascular death after transient ischemic attack and ischemic stroke: a systematic review and
meta-analysis. Stroke; a journal of cerebral circulation. 2005;36(12):2748-2755.
National Center for Health Statistics. National Health Interview Survey, 2009-2011.
ftp://ftp.cdc.gov/pub/Health_Statistics/NCHS/Datasets/NHIS/. Accessed June 12, 2012.
Bergmann MM, Byers T, Freedman DS, Mokdad AH. Validity of self-reported diagnoses leading
to hospitalization: a comparison of self-reports with hospital records in a prospective study of
American adults. American journal of epidemiology. 1998;147(10):969-977.
©Institute for Clinical and Economic Review, 2015
Page 79
164.
165.
Ford ES, Giles WH. Changes in prevalence of nonfatal coronary heart disease in the United
States from 1971-1994. Ethnicity and Disease. 2003;13:85-93.
Gross R, Bentur N, Elhayany A, Sherf M, Epstein L. The validity of self-reports on chronic disease:
characteristics of underreporters and implications for the planning of services. Public Health
Reviews. 1996;24:167-182.
©Institute for Clinical and Economic Review, 2015
Page 80
APPENDICES
©Institute for Clinical and Economic Review, 2015
Page 81
A1. Search Strategies
PubMed
((((((alirocumab) OR evolocumab) OR pcsk9 inhibitor) OR pcsk9 antibody) OR amg 145) OR
regn727) OR sar236553
Cochrane
((((((alirocumab) OR evolocumab) OR pcsk9 inhibitor) OR pcsk9 antibody) OR amg 145) OR
regn727) OR sar236553
©Institute for Clinical and Economic Review, 2015
Page 82
Figure 1. Selection of Studies for Inclusion in Review
442 potentially relevant
references screened
101 duplicate citations excluded
249 abstracts for assessment
92 excluded: reviews, other PCSK9
inhibitors
154 references excluded
95 references for full text
review
(Editorials, reviews, no clinical
outcomes)
70 references excluded: no primary
data, multiple publications, pooled
data, phase 1 trials
25 unique trials
©Institute for Clinical and Economic Review, 2015
Page 83
A2. Clinical Guidelines
For the purposes of this review, this section outlines available clinical guidelines for the
management of high cholesterol. It is important to note that that the use of target or goal lipid
levels is no longer universal across all guideline statements. As noted in the Background section,
this has been the subject of debate in many policy settings and the academic literature. At the time
of review, existing guidelines did not include reference to PCSK9 inhibitors for treatment of
cholesterol. Websites for this review were accessed on August 17, 2015.
Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on
Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment
Panel III) Final Report 2001
http://circ.ahajournals.org/content/106/25/3143.full.pdf
In 2001 the National Cholesterol Education Program (NCEP) produced a report updated
recommendations for cholesterol testing and management (ATP III). Many of the guidelines
included in this section build upon this report. ATP III targets the clinical approach to prevention of
coronary heart disease (CHD) and identifies low-density lipoprotein (LDL-C) as the primary target of
cholesterol-lowering therapy. Statins are noted as the primary option for attainment of the LDL-C
goal in higher-risk individuals as they are most effective, well tolerated and easy to administer.
Combination therapy with other agents (e.g., bile acid sequestrants) may be needed to provide
additional reduction of LDL-C, to achieve the goal for non-HDL cholesterol, to treat severe
hypertriglyceridemia, and if it seems advisable, to raise HDL cholesterol levels.
The report Identifies LDL-C of <100 mg/dL as optimal and notes that prospective epidemiological
studies show that when LDL–C levels are below 100 mg/dL an individual’s CHD risk likewise is low,
even in the presence of other risk factors. However, in 2004, the National Cholesterol Education
Program (NCEP) Adult Treatment Program (ATP) III updated its guidelines to include an "optional"
LDL-C goal less than 70 mg/dL for patients at very high risk. The 2004 NCEP ATP III update further
indicated that it is always prudent to initiate therapy at a level sufficient to achieve a 30% to 40%
LDL-C reduction.i
©Institute for Clinical and Economic Review, 2015
Page 84
Guidelines from Clinical Societies
2013 ACC/AHA Guideline on the Treatment of Blood Cholesterol to Reduce Atherosclerotic
Cardiovascular Risk in Adults
http://circ.ahajournals.org/content/early/2013/11/11/01.cir.0000437738.63853.7a
The previous ACC/AHA guideline recommended that clinicians treat patients to a specific LDL-C
target by stratifying patients according to their cardiovascular risk, assigning an LDL-C target, and
titrating medication use to reach that target. The updated guideline does not contain
recommendations for or against specific LDL-C or non–HDL-C targets for the primary or secondary
prevention of ASCVD. The Expert Panel was unable to find sufficient RCT evidence to support
continued use of specific LDL-C or non–HDL-C treatment targets.
The guideline describes 4 statin benefit groups that focus efforts to reduce ASCVD events in
secondary and primary prevention. The appropriate intensity of statin therapy should be used to
reduce ASCVD risk in those most likely to benefit.
Primary Prevention
Adults ≥21 years of age with primary, severe elevations of LDL–C (≥190 mg/dL) have a high lifetime
risk for ASCVD events and should receive high-intensity statin therapy if they have not already been
diagnosed and treated before this age. For individuals with LDL-C 70–189 mg/dL moderateintensity statin therapy should be initiated or continued for adults 40 to 75 years of age with
diabetes. High-intensity statin therapy is reasonable for adults 40 to 75 years of age with diabetes
with a ≥7.5% estimated 10-year ASCVD risk unless contraindicated. Adults 40 to 75 years of age with
LDL-C 70–189 mg/dL, without clinical ASCVD or diabetes, and with an estimated 10-year ASCVD risk
≥7.5% should be treated with moderate- to high-intensity statin therapy. It is reasonable to offer
treatment with a moderate-intensity statin to adults 40 to 75 years of age, with LDL-C 70–189
mg/dL, without clinical ASCVD or diabetes, and with an estimated 10-year ASCVD risk of 5% to
<7.5%.
Secondary Prevention
The guideline recommends that secondary prevention high-intensity statin therapy should be
initiated or continued as first-line therapy in women and men ≤75 years of age who have clinical
ASCVD, unless contraindicated. Moderate-intensity statin should be used as the second option if
tolerated in individuals with clinical ASCVD in whom high-intensity statin therapy would otherwise
be used, when high-intensity statin therapy is contraindicated or when characteristics predisposing
to statin-associated adverse effects are present. Individuals with clinical ASCVD >75 years of age
should be evaluated for potential ASCVD risk-reduction benefits and for adverse effects and drug
©Institute for Clinical and Economic Review, 2015
Page 85
interactions; patient preferences should also be considered when initiating a moderate- or highintensity statin.
Non statin therapies, as compared to statin therapy, do not provide acceptable ASCVD riskreduction benefits relative to their potential for adverse effects in the routine prevention of ASCVD.
If considering non-statin therapies, the guideline recommends that adherence to lifestyle and to
statin therapy re-emphasized before the addition of a non-statin drug. High-risk patients who have
a less-than-anticipated response to statins, who are unable to tolerate a less-than-recommended
intensity of a statin, or who are completely statin intolerant may require a non-statin cholesterollowering therapy. Non statin therapies referenced in the guideline include niacin, bile acid
sequestrants, cholesterol-absorption inhibitors, fibrates, and omega-3 fatty acids.
The American Association of Clinical Endocrinologists’ (AACE) Guidelines for Management of
Dyslipidemia and Prevention of Atherosclerosis (March and April 2012)
https://www.aace.com/sites/default/files/LipidGuidelines.pdf
For patients with average or elevated LDL-C, the AACE recommends aggressive lipid-modifying
therapy to lower LDL-C to less than 100 mg/dL.
The AACE identifies statins as the drug of choice for LDL-C reduction due to the efficacy and safety
profile of the drug class. Other pharmacologic therapy referenced in the guidelines include fibrates,
niacin, bile acid sequestrants, cholesterol absorption inhibitors (ezetimibe), and combination
therapy.
Cholesterol Absorption Inhibitors (ezetimibe) are noted as an effective monotherapy in reducing
LDL-C and apo B or as part of combination therapy with statins. Current research indicates that
combining statins with ezetimibe may yield enhanced benefits by furthering the impact of statins
on triglycerides and HDL-C, but it is uncertain whether ezetimibe has a direct benefit on reducing
cardiovascular events.
In adults of both sexes, AACE recommends a target LDL-C concentration less than 100 mg/dL and
less than 70 mg/dL in patients at very high risk, including those with one or more additional risk
factors, such as established CVD or CAD. For patients with diabetes mellitus, AACE recommends an
LDL-C goal of less than 100 mg/dL. Lipid goals for all patients should be personalized by levels of risk
and suggest that there is no threshold below which LDL-C lowering ceases to be effective, stating
that reducing lipids to levels even below recommended targets may be beneficial for certain
patients (e.g., those with metabolic syndrome).
©Institute for Clinical and Economic Review, 2015
Page 86
AACE recommends aggressive therapy for patients undergoing coronary artery bypass graft,
patients with acute coronary syndrome, and certain healthy and functional older patients at high
risk who may be appropriate candidates for aggressive therapy.
European Society of Cardiology (ESC)/European Atherosclerosis Society (EAS) Guidelines for
the Management of Dyslipidaemias
http://www.escardio.org/static_file/Escardio/Guidelines/publications/DYSLIPguidelinesdyslipidemias-FT.pdf
For pharmacologic treatment of hypercholesterolemia, ESC/EAS recommends that physicians
prescribe statins up to the highest recommended dose, or to the highest dose tolerated by the
patient. In patients who do not reach LDL-C targets with statin therapy alone, a combination of a
statin and a bile acid sequestrant, nicotinic acid, or cholesterol absorption inhibitor may be
considered. If a patient is statin intolerant, bile acid sequestrants or nicotinic acid should be
considered. A cholesterol absorption inhibitor, either alone or combined with a bile acid
sequestrant or nicotinic acid, may also be considered.
For patients with HeFH, a high dose statin is recommended. If needed, a combination of statin and
cholesterol absorption inhibitor and/or bile acid sequestrant and is recommended. Treatment
should be aimed at reaching LDL-C goals of <100 mg/dL in high risk subjects. A goal of <70 mg/dL is
recommended for high risk subjects with CVD. If targets cannot be reached, maximal reduction of
LDL-C using appropriate drug combinations is recommended.
National Lipid Association Recommendations for Patient-Centered Management of
Dyslipidemia (September 2014)
http://www.lipidjournal.com/article/S1933-2874(14)00274-8/pdf
The National Lipid Association recommends that for patients in whom lipid-lowering drug therapy is
indicated, statin treatment is the primary modality for reducing ASCVD risk. The recommendations
emphasize that non-HDL-C is a better primary target for modification than LDL-C, and is considered
to be a co-target with LDL-C.
These recommendations note that patients experiencing statin intolerance may improve if switched
to a different statin. Alternative strategies include limiting the daily dosage and modified regimens
such as every other day or once weekly dosing with statins. In some patients, it may be possible to
switch to an alternative concomitant therapy to enhance statin tolerance. For patients who cannot
tolerate a statin with the previously discussed strategies, a non-statin drug alone or in combination
with another cholesterol lowering agent may be considered.
©Institute for Clinical and Economic Review, 2015
Page 87
Non-statin drug classes for lipid management include cholesterol absorption inhibitors (ezetimibe),
bile acid sequestrants, fibric acids, long-chain omega-3 fatty acid concentrates, and nicotinic acid.
The recommendation references two additional classes of medications available with more limited
indications for the treatment of patients with homozygous familial hypercholesterolemia (FH): an
antisense oligonucleotide that targets the messenger RNA for apo B, and a microsomal triglyceride
transfer protein inhibitor.
Patients with severe hypercholesterolemia phenotype (LDL-C >190 mg/dL) are at increased lifetime
risk for ASCVD, particularly premature ASCVD, and may not be able to achieve goal cholesterol
levels even with combination drug therapy. When this is the case, an alternative goal should be to
lower LCL-C levels by at least 50%. The recommendation references PCSK9 inhibitors (under
investigation at the time of publication) as having the potential to make attainment of goal
cholesterol levels practical for a greater fraction of patients with severe hypercholesterolemia.
The LDL-C goals for therapy for patients at risk for CAD are LDL-C of <100 mg/dL and LDL-C of <70
mg/dL for all very high risk patients.
Lipid modification: cardiovascular risk assessment and the modification of blood lipids for the
primary and secondary prevention of cardiovascular disease. National Institute for Health
Care and Excellence (NICE) (July 2014)
http://www.nice.org.uk/guidance/cg181/chapter/1-recommendations#identifying-and-assessingcardiovascular-disease-cvd-risk-2
Considerations for statin use should follow a formal assessment of cardiovascular risk, except in
patients with type 1 diabetes, familial hypercholesterolemia, pre-existing CVD, or eGFR rates less
than 60/ml/min/1.73m2. These groups have high cardiovascular risk and do not require assessment.
For primary prevention, patients should first be offered support for lifestyle modification. If lifestyle
interventions are ineffective or unsuited for the patient, statins should be offered to patients with a
10% or greater 10-year risk of developing CVD based on the QRISK2 assessment tool. Dosing should
begin with 20mg of atorvastatin. Patients with type 1 diabetes should be offered 20 mg of
atorvastatin if they are 40 years or older, have had diabetes for at least 10 years, have established
nephropathy, or have other CVD risk factors. Patients age 85 or older should be considered for
statin therapy for prevention of myocardial infarction.
For secondary prevention, patients should receive 80 mg of atorvastatin. In the case of potential
drug interactions or high risk of adverse reactions, or based on patient preference, a lower dose
may be used. If a person is unable to tolerate a high-intensity statin, treat with the maximally
tolerated dose. If patients report adverse effects, physicians should try reducing the dose, stopping
and restarting the statin to see if symptoms are related to use, or change the statin to a lower
intensity group.
©Institute for Clinical and Economic Review, 2015
Page 88
The European Atherosclerosis Society consensus statement on treating familial
hypercholesterolemia (August 2013)
http://eurheartj.oxfordjournals.org/content/early/2013/08/15/eurheartj.eht273.full
In August of 2013 the European Atherosclerosis Society released a consensus statement, providing
guidance for screening and treatment of familial hypercholesterolemia, in order to prevent
coronary heart disease (CHD). The statement recommends initiation of cholesterol-lowering drugs
immediately at diagnosis in adults along with lifestyle management. The statement ranks specific
priorities for pharmacotherapy beginning with maximal potent statin dose, followed by ezetimibe,
then bile acid-binding resins, and finally lipoprotein apheresis in homozygotes and in treatmentresistant heterozygotes with CHD. The consensus statement recommends treating to specific LDL-C
targets. For adults LDL-C targets are LDL-C of <100 mg/dL and LDL-C of <70 for adults with known
CHD or diabetes. The statement advises a clinical assessment of efficacy and safety 4–6 weeks after
initiating treatment.
©Institute for Clinical and Economic Review, 2015
Page 89
A3. Detailed Coverage Policies
Medicaid
Rosuvastatin (Crestor®)
Of the six New England states, three cover rosuvastatin (Crestor®) without restriction, but quantity
limits may apply. The remaining three states require prior authorization or step therapy.
Massachusetts requires that patients have tried atorvastatin at a dose of at least 80 mg per day (or
a similar statin with equivalent potency) and have not had adequate reduction in LDL-C, or have a
contraindication to atorvastatin. Rhode Island requires prior authorization and requires that
patients first try atorvastatin. New Hampshire lists Crestor as a non-preferred agent, which typically
signals non-coverage.
Ezetimibe/simvastatin (Vytorin®)
With the exception of Maine, every state in New England limits coverage of ezetimibe/simvastatin
(Vytorin®). Coverage in Massachusetts is limited to those for whom a prior regimen of atorvastatin
at a dose of at least 80 mg/day (or a similar statin with equivalent potency) has failed to achieve
adequate reduction in LDL-C , or who havea contraindication to atorvastatin. Vermont utilizes prior
authorization and requires a prior regimen of atorvastatin or Crestor. Rhode Island also requires
prior authorization. New Hampshire lists Vytorin as a non-preferred agent, and Connecticut does
not list Vytorin on its formulary.
Ezetimibe (Zetia®)
A majority of New England states utilize prior authorization criteria for ezetimibe (Zetia®). In Maine,
Massachusetts, and Vermont, prior authorization policies require a prior regimen of a statin with
inadequate response or the presence of a contraindication to statins. Rhode Island also utilizes prior
authorization, though the requirements are not publicly available. New Hampshire requires prior
regimen of at least 2 high-potency statins or combination products. Connecticut does not list Zetia®
on its formulary.
Regional Private Payers
Rosuvastatin (Crestor®)
Major regional private payers in New England generally cover Crestor as a tier 2 or tier 3 drug.
Blue Cross Blue Shield of Massachusetts (BCBSMA), Neighborhood Health Plan of Rhode Island
(NHPRI), and ConnectiCare all require step therapy, while BCBSMA and ConnectiCare also
©Institute for Clinical and Economic Review, 2015
Page 90
impose quantity limits. Tufts Health Plan (THP) requires prior authorization. Blue Cross Blue
Shield of Vermont (BCBSVT) covers Crestor without restriction.
Ezetimibe/simvastatin (Vytorin®)
Most regional plans cover Vytorin as a tier 3 drug, with the exception of THP which covers it as
tier 2. NHPRI and ConnectiCare require step therapy, as does Harvard Pilgrim Health Care
(HPHC) for higher dose formulations. THP requires prior authorization. BCBSMA and
ConnectiCare also use quantity limits. BCBSMA lists Vytorin as a non-covered medication, but
physicians may request coverage if no other alternative is suitable for treatment of a patient’s
condition. Step therapy and quantity limits apply.
Ezetimibe (Zetia®)
Regional plans cover Zetia as a tier 2 or tier 3 medication. BCBSMA and HPHC require step therapy,
and NHPRI requires prior authorization. ConnectiCare imposes quantity limits.
National Private Payers/Pharmacy Benefit Managers
Rosuvastatin (Crestor®)
On a national level, private payers generally list Crestor as a tier 2 drug and apply quantity limits.
Cigna also applies prior authorization and step therapy for the 5mg and 10mg formulations, and
offers a deductible exemption under its preventative drug benefit.
Ezetimibe/simvastatin (Vytorin®)
Most national private payers apply similar criteria to Vytorin, listing it as a tier 2, 3, or 4 drug,
depending on the plan. Aetna and Cigna require step therapy. Aetna also applies quantity limits, as
do Humana and United Healthcare. Cigna requires prior authorization.
Ezetimibe (Zetia®)
Zetia is generally covered as a tier 2 drug, except in the case of United Healthcare, which covers it as
a tier 3 or tier 4 drug. Aetna, Cigna, Humana, and United Healthcare apply quantity limits. Cigna
offers a deductible exemption through its preventative drug benefit.
©Institute for Clinical and Economic Review, 2015
Page 91
A4. Previous Systematic Reviews and
Technology Assessments
We did not identify any prior technology assessments addressing PCSK9 inhibitor therapy. There are
two published systematic reviews and meta-analyses that were cited as part of the evidence
review. Both reviews identified the same 25 studies included in this New England CEPAC
assessment. They are briefly summarized below.
Navarese EP, Kolodziejczak M, Schulze V, et al. Effects of Proprotein Convertase Subtilisin/Kexin
Type 9 Antibodies in Adults With Hypercholesterolemia: A Systematic Review and Meta-analysis.
Ann Intern Med. 2015. PMID: 25915661.
The investigators performed a high quality systematic review that included phase 2 or 3 randomized
trials of PCSK9 antibodies, but excluded the OSLER study (combined OSLER 1 and OSLER 2) from
their analyses because all of the patients in the OSLER study were included in the 24 earlier studies
(n=10,159). The investigators combined the results for both alirocumab and evolocumab. They
found that treatment with PCSK9 inhibitors markedly lowered LDL-C levels and reduced all-cause
mortality that they describe as “an important preliminary signal of survival benefit compared with
no anti PCSK9 treatment.” They concluded that PCSK9 antibodies seem safe and effect for adults
with dyslipidemia. They note that their conclusions are limited by the rare clinical outcome data and
the fact that study level data rather than patient level data were used for their analyses.
Zhang XL, Zhu QQ, Zhu L, et al. Safety and efficacy of anti-PCSK9 antibodies: a meta-analysis of 25
randomized, controlled trials. BMC Med. 2015;13:123. PMID: 26099511.
The authors combined results from the 25 trials of 12,200 patients, including patients from the
OSLER 1 study twice. They focused much more heavily on evolocumab than alirocumab. They found
that evolocumab every two or four weeks reduced LDL-C by 55% versus placebo and 36% versus
ezetimibe. Alirocumab 50-150 mg every two weeks reduced LDL-C by 53% versus placebo and 30%
versus ezetimibe. Overall they found no significant differences in rates of common adverse events
with placebo or ezetimibe controls. Alirocumab, but not evolocumab was associated with a reduced
risk of death (p=0.04). Evolocumab was associated with a reduced rate of abnormal liver function
(p=0.03). They concluded that both drugs were safe and well tolerated with favorable changes in all
lipid parameter. They end stating that they “await the results of ongoing trials evaluating their
effects on CVD events.”
©Institute for Clinical and Economic Review, 2015
Page 92
A5. Ongoing Studies
Title/Trial Sponsor
Study Design
Comparators
Patient Population
Primary Outcomes
Estimated
Completion Date
Alirocumab
ODYSSEY OUTCOMES
RCT
Alirocumab
Sanofi / Regeneron
N = 18,000
Placebo
NCT01663402
5-6 year FU
• Post-acute coronary syndrome
• Age 40 years
• ACS in past 52 weeks
Combination of CHD
death, MI, stroke,
unstable angina.
Dec 2017
Combination of CHD
death, MI, stroke,
unstable angina, coronary
re-vascularization
Feb 2018
• LDL-C ≥ 70 mg/dL
Evolocumab
FOURIER
RCT
Evolocumab
• History of CVD event at high risk for
recurrence
AMGEN
N = 22,500
Placebo
• Age 40-85 years
NCT01764633
5 year FU
©Institute for Clinical and Economic Review, 2015
• LDL-C ≥ 70 mg/dL or non-HDL ≥ 100
mg/dL
Page 93
A6. Comparative Clinical Effectiveness Appendix
A6 Table 1: Overview of studies
Reference
Study
Phase
Alirocumab
Heterozygous (HeFH) or mixed familial hyperlipidemia (FH)
Stein 2012
–
2
–
ODYSSEY FH I
3
–
ODYSSEY FH II
3
–
ODYSSEY HIGH FH
3
Robinson 2015
ODYSSEY Long Term
3
McKenney 2012
–
2
Roth 2012
–
2
Hypercholesterolemia (HC)
–
ODYSSEY ALTERNATIVE
Kereiakas 2015
ODYSSEY COMBO I
Cannon 2015
ODYSSEY COMBO II
Roth 2014
ODYSSEY MONO
Bays 2015
ODYSSEY OPTIONS I
–
ODYSSEY OPTIONS II
3
3
3
3
3
3
N
FU, weeks Treatment
Control
Population
Statin
Therapy
Age, years
Sex, %F
Prior CVD,
%
DM, %
77
486
249
107
2341
183
92
12
78
78
52-78
78
12
8
Alirocumab 150 mg q 2 weeks
Alirocumab 75-150 mg q 2 weeks
Alirocumab 75-150 mg q 2 weeks
Alirocumab 150 mg q 2 weeks
Alirocumab 150 mg q 2 weeks
Alirocumab 150 mg q 2 weeks
Alirocumab 150 mg q 2 weeks
Placebo
Placebo
Placebo
Placebo
Placebo
Placebo
Placebo
HeFH
HeFH
HeFH
HeFH
HeFH / HC
FH
FH
Intensive
Intensive
Intensive
Intensive
Intensive
Intensive
Intensive
53
52
53
52
61
57
57
39
55
47
47
38
53
60
42
46
35
50
69
6
3
4
12
4
14
35
12
15
251
316
720
103
205
204
24
52
104
24
24
24
Alirocumab 75-150 mg q 2 weeks
Alirocumab 75-150 mg q 2 weeks
Alirocumab 75-150 mg q 2 weeks
Alirocumab 75 mg q 2 weeks
Alirocumab 75-150 mg q 2 weeks
Alirocumab 75-150 mg q 2 weeks
Ezetimibe 10 mg
Placebo
Ezetimibe 10 mg
Ezetimibe 10 mg
Ezetimibe 10 mg
Ezetimibe 10 mg
HC
HC
HC
HC
HC
HC
None
Intensive
Intensive
None
Mixed
Mixed
63
63
61
60
66
60
45
34
74
47
36
43
47
78
81
NR
NR
NR
24
43
31
4
NR
NR
49
12
Evolocumab 420 mg q 4 weeks
Placebo
HoFH
Intensive
31
49
43
NR
167
329
614
12
12
12
Evolocumab 420 mg q 4 weeks
Placebo
Evolocumab 140 mg q 2 weeks and Evolocumab 420 mg q 4 weeks Placebo
Evolocumab 140 mg q 2 weeks and Evolocumab 420 mg q 4 weeks Placebo and ezetimibe 10 mg
HeFH
HeFH
FH
Intensive
Intensive
None
50
51
54
47
42
69
21
62
0
NR
31
0.1
157
307
629
406
307
1896
901
12
12
12
12
12
12
52
Evolocumab 420 mg q 4 weeks
Evolocumab 140 mg q 2 weeks and Evolocumab 420 mg q 4 weeks
Evolocumab 140 mg q 2 weeks and Evolocumab 420 mg q 4 weeks
Evolocumab 140 mg q 2 weeks and Evolocumab 420 mg q 4 weeks
Evolocumab 140 mg q 2 weeks and Evolocumab 420 mg q 4 weeks
Evolocumab 140 mg q 2 weeks and Evolocumab 420 mg q 4 weeks
Evolocumab 420 mg q 4 weeks
HC, SI
HC, SI
HC
HC
HC
HC
HC
Nonintensive
Nonintensive
Intensive
None
Intensive
Mixed
Mixed
62
62
62
51
62
60
57
64
46
51
66
37
46
52
13
NR
30
0
25
23
15
NR
20
16
0.2
38
16
12
Mixed
Mixed
58
51
20
13
5334
Evolocumab
Homozygous familial hyperlipidemia (HoFH)
Raal 2015
TESLA Part B
3
Heterozygous (HeFH) or mixed familial hyperlipidemia (FH)
Raal 2012
RUTHERFORD
2
Raal 2015
RUTHERFORD 2
3
Koren 2014
MENDEL 2
3
Hypercholesterolemia (HC)
Sullivan 2012
GAUSS
Stroes 2014
GAUSS 2
Giuliano 2012
LAPLACE TIMI 57
Koren 2012
MENDEL
Hirayama 2014
YUKAWA
Robinson 2014
LAPLACE 2
Blom 2014
DESCARTES
2
3
2
2
2
3
3
Long-term FU of phase 2 and 3 studies - all patients already described above
Sabatine 2014
OSLER
–
4465
52
Ezetimibe 10 mg
Ezetimibe 10 mg
Placebo
Placebo and ezetimibe 10 mg
Placebo
Placebo and ezetimibe 10 mg
Placebo
Evolocumab 140 mg q 2 weeks and Evolocumab 420 mg q 4 weeks Standard therapy
©Institute for Clinical and Economic Review, 2015
Page 94
A6 Table 2: Quality of the studies
Reference
Study
Alirocumab
Heterozygous (HeFH) or mixed familial hyperlipidemia (FH)
Stein 2012
–
–
ODYSSEY FH I
–
ODYSSEY FH II
–
ODYSSEY HIGH FH
Robinson 2015
ODYSSEY Long Term
McKenney 2012
–
Roth 2012
–
Adequate
Allocation
randomization concealment
Patent
blinding
Outcome
adjudication Completeness Intention to
Staff blinding
blinding
of follow-up treat analysis
Incomplete
data
addressed
Selective
outcome
reporting
Industry
funding
Free from
other bias
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
No
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Hypercholesterolemia (HC)
–
ODYSSEY ALTERNATIVE
Kereiakas 2015
ODYSSEY COMBO I
Cannon 2015
ODYSSEY COMBO II
Roth 2014
ODYSSEY MONO
Bays 2015
ODYSSEY OPTIONS I
–
ODYSSEY OPTIONS II
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Evolocumab
Homozygous familial hyperlipidemia (HoFH)
Raal 2015
TESLA Part B
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Heterozygous (HeFH) or mixed familial hyperlipidemia (FH)
Raal 2012
RUTHERFORD
Raal 2015
RUTHERFORD 2
Koren 2014
MENDEL 2
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
No
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
Yes
Yes
Yes
No
Yes
Yes
Hypercholesterolemia (HC)
Sullivan 2012
GAUSS
Stroes 2014
GAUSS 2
Giuliano 2012
LAPLACE TIMI 57
Koren 2012
MENDEL
Hirayama 2014
YUKAWA
Robinson 2014
LAPLACE 2
Blom 2014
DESCARTES
Long-term FU of phase 2 and 3 studies - all patients already described above
Sabatine 2014
OSLER 1 and 2
Yes
©Institute for Clinical and Economic Review, 2015
Page 95
A6 Table 3: LDL outcomes
Q2W
Reference
Study
LDL < 70 mg/dL, % reduction in
%
LDL vs placebo
Q4W
Baseline LDL
Final LDL
Alirocumab
Placebo
Alirocumab
Placebo
Alirocumab
Placebo
Alirocumab
Placebo
Alirocumab
Placebo
Alirocumab
Placebo
Alirocumab
Placebo
147
151
140
139
140
139
196
201
123
122
124
130
127
121
50
136
70
145
70
145
113
188
48
119
34
127
54
104
81
0
77
7
77
7
32
3
79
8
100
3
90
17
Alirocumab
Ezetimibe
Alirocumab
Placebo
Alirocumab
Ezetimibe
Alirocumab
Ezetimibe
Alirocumab
Ezetimibe
Alirocumab
Ezetimibe
191
194
100
106
109
105
141
138
110
100
113
110
92
157
52
102
52
82
87
121
52
78
43
69
42
4
5
9
77
46
NR
NR
78
52
68
49
Evolocumab
Placebo
356
336
282
356
NR
NR
30.9
Evolocumab
Placebo
Evolocumab
Placebo
Evolocumab
Ezetimibe
Placebo
151
162
158
151
143
144
142
70
162
68
153
62
117
141
65
0
66
2
69
1.4
0.7
56.4
Evolocumab
Ezetimibe
Evolocumab
Ezetimibe
Evolocumab
Placebo
Evolocumab
Ezetimibe
Placebo
Evolocumab
Placebo
Evolocumab
Ezetimibe
Evolocumab
Evolocumab
Placebo
204
183
192
195
120
124
139
143
145
139
143
110
109
108
104
104
99
154
90
162
63
122
72
122
147
41
139
44
89
112
51
108
NR
0
44
1
83
1
NR
NR
NR
86
0
89
37
12
82
6.4
Long-term FU of phase 2 and 3 studies - all patients already described above
Sabatine 2014
OSLER
Evolocumab
Placebo
2976
1489
120
121
73.6
3.8
–
ODYSSEY FH I
–
ODYSSEY FH II
–
ODYSSEY HIGH FH
Robinson 2015
ODYSSEY Long Term
McKenney 2012
–
Roth 2012
–
Hypercholesterolemia (HC)
–
ODYSSEY ALTERNATIVE
Kereiakas 2015
ODYSSEY COMBO I
Cannon 2015
ODYSSEY COMBO II
Roth 2014
ODYSSEY MONO
Bays 2015
ODYSSEY OPTIONS I
–
ODYSSEY OPTIONS II
Evolocumab
Homozygous familial hyperlipidemia (HoFH)
Raal 2015
TESLA Part B
Heterozygous (HeFH) or mixed familial hyperlipidemia (FH)
Raal 2012
RUTHERFORD
Raal 2015
RUTHERFORD 2
Koren 2014
MENDEL 2
Hypercholesterolemia (HC)
Sullivan 2012
GAUSS
Stroes 2014
GAUSS 2
Giuliano 2012
LAPLACE TIMI 57
Koren 2012
MENDEL
Hirayama 2014
YUKAWA
Robinson 2014
LAPLACE 2
Blom 2014
DESCARTES
©Institute for Clinical and Economic Review, 2015
57.2
Q4W
% reduction in % reduction in % reduction in
LDL vs placebo LDL vs ezetimibe LDL vs ezetimibe
Intervention
Alirocumab
Heterozygous (HeFH) or mixed familial hyperlipidemia (FH)
Stein 2012
–
Q2W
23.5
57.9
51.4
39.1
61.9
53.5
40.4
48.9
30.4
45.9
29.8
31.6
27.2
30.5
59.2
61.3
49.6
52.8
35.8
47.3
66.1
50.3
47.2
52.5
68.6
63.9
70.9
61.9
34
35.9
38.1
37.6
36.7
34.1
43.4
40
58.4
0.9
1.3
Page 96
A6 Table 4: Clinical outcomes / Major adverse cardiac outcomes (MACE)
Reference
Study
Alirocumab
Heterozygous (HeFH) or mixed familial hyperlipidemia (FH)
Stein 2012
–
–
ODYSSEY FH I and II
–
ODYSSEY FH II
–
ODYSSEY HIGH FH
Robinson 2015
ODYSSEY Long Term
McKenney 2012
–
Roth 2012
–
Hypercholesterolemia (HC)
–
ODYSSEY ALTERNATIVE
Kereiakas 2015
ODYSSEY COMBO I
Cannon 2015
ODYSSEY COMBO II
Roth 2014
ODYSSEY MONO
Bays 2015
ODYSSEY OPTIONS I
–
ODYSSEY OPTIONS II
Evolocumab
Homozygous familial hyperlipidemia (HoFH)
Raal 2015
TESLA Part B
Heterozygous (HeFH) or mixed familial hyperlipidemia (FH)
Raal 2012
RUTHERFORD
Raal 2015
RUTHERFORD 2
Koren 2014
MENDEL 2
Hypercholesterolemia (HC)
Sullivan 2012
GAUSS
Stroes 2014
GAUSS 2
Giuliano 2012
LAPLACE TIMI 57
Koren 2012
MENDEL
Hirayama 2014
YUKAWA
Robinson 2014
LAPLACE 2
Blom 2014
DESCARTES
Intervention
N
Death
CVD death
MI
Stroke
Angina
MACE
Alirocumab
Placebo
Alirocumab
Placebo
Alirocumab
Placebo
Alirocumab
Placebo
Alirocumab
Placebo
Alirocumab
Placebo
Alirocumab
Placebo
16
15
490
245
167
82
72
35
1553
788
31
31
30
31
0
0
4
0
0
0
0
0
11
8
0
0
0
0
0
0
2
0
0
0
0
0
6
6
0
0
0
0
0
0
2
0
0
1
4
0
13
17
0
0
0
0
0
0
1
0
0
0
0
0
10
3
0
0
0
0
0
0
1
0
0
0
0
0
0
1
0
0
0
0
0
0
5
0
0
1
4
0
27
25
0
0
0
0
Alirocumab
Ezetimibe
Alirocumab
Placebo
Alirocumab
Ezetimibe
Alirocumab
Ezetimibe
Alirocumab
Ezetimibe
Alirocumab
Ezetimibe
126
125
209
107
479
241
52
51
104
102
103
101
0
0
3
2
2
4
0
0
0
0
0
1
0
0
2
1
2
2
0
0
0
0
0
0
1
0
2
1
12
3
0
0
0
0
0
1
0
0
2
0
1
1
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
0
5
2
16
5
0
0
0
0
0
1
Evolocumab
Placebo
33
16
0
0
0
0
0
0
0
0
0
0
0
0
Evolocumab
Placebo
Evolocumab
Placebo
Evolocumab
Ezetimibe
Placebo
56
56
221
110
306
308
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Evolocumab
Ezetimibe
Evolocumab
Ezetimibe
Evolocumab
Placebo
Evolocumab
Ezetimibe
Placebo
Evolocumab
Placebo
Evolocumab
Ezetimibe
32
33
205
102
158
157
90
135
0
0
0
0
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
105
102
1117
779
0
0
0
1
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
Evolocumab
Placebo
599
302
2
0
2
0
1
0
1
0
3
0
2976
1489
4
6
4
3
9
5
3
3
28
30
Long-term FU of phase 2 and 3 studies - all patients already described above
Sabatine 2014
OSLER
Evolocumab
Placebo
©Institute for Clinical and Economic Review, 2015
3
2
Page 97
A6 Table 5: Adverse outcomes
Reference
Study
Alirocumab
Heterozygous (HeFH) or mixed familial hyperlipidemia (FH)
Stein 2012
–
–
ODYSSEY FH I and II
–
ODYSSEY FH II
–
ODYSSEY HIGH FH
Robinson 2015
ODYSSEY Long Term
McKenney 2012
–
Roth 2012
–
Hypercholesterolemia (HC)
–
ODYSSEY ALTERNATIVE
Kereiakas 2015
ODYSSEY COMBO I
Cannon 2015
ODYSSEY COMBO II
Roth 2014
ODYSSEY MONO
Bays 2015
ODYSSEY OPTIONS I
–
ODYSSEY OPTIONS II
Evolocumab
Homozygous familial hyperlipidemia (HoFH)
Raal 2015
TESLA Part B
Heterozygous (HeFH) or mixed familial hyperlipidemia (FH)
Raal 2012
RUTHERFORD
Raal 2015
RUTHERFORD 2
Koren 2014
MENDEL 2
Hypercholesterolemia (HC)
Sullivan 2012
GAUSS
Stroes 2014
GAUSS 2
Giuliano 2012
LAPLACE TIMI 57
Koren 2012
MENDEL
Hirayama 2014
YUKAWA
Robinson 2014
LAPLACE 2
Blom 2014
DESCARTES
Discontinue
CK
due to AE elevation
ALT
elevation
Intervention
N
SAE
Alirocumab
Placebo
Alirocumab
Placebo
Alirocumab
Placebo
Alirocumab
Placebo
Alirocumab
Placebo
Alirocumab
Placebo
Alirocumab
Placebo
16
15
490
245
167
82
72
35
1553
788
31
31
30
31
0
1
49
22
0
0
0
0
17
15
0
0
8
4
290
154
0
1
1
0
3
1
111
46
1
0
1
4
2
0
56
38
0
1
0
1
3
1
28
16
0
0
0
0
Alirocumab
Ezetimibe
Alirocumab
Placebo
Alirocumab
Ezetimibe
Alirocumab
Ezetimibe
Alirocumab
Ezetimibe
Alirocumab
Ezetimibe
126
125
209
107
479
241
52
51
104
102
103
101
12
10
26
14
90
43
1
1
4
7
6
8
23
31
13
8
36
13
5
4
7
4
5
8
3
2
4
5
13
6
0
1
3
1
0
3
0
0
Evolocumab
Placebo
33
16
0
0
0
0
1
1
2
1
Evolocumab
Placebo
Evolocumab
Placebo
Evolocumab
Ezetimibe
Placebo
56
56
221
110
306
308
2
0
7
5
4
2
1
1
0
0
7
11
2
0
0
2
2
2
1
0
0
0
3
6
3
6
Evolocumab
Ezetimibe
Evolocumab
Ezetimibe
Evolocumab
Placebo
Evolocumab
Ezetimibe
Placebo
Evolocumab
Placebo
Evolocumab
Ezetimibe
23
33
205
102
158
157
90
135
0
0
6
4
6
4
1
0
1
2
17
13
2
0
0
2
0
1
0
2
2
0
0
1
0
0
0
0
0
1
0
0
6
1
16
18
1
2
2
1
105
102
1117
779
3
0
23
15
3
0
21
16
0
1
1
2
0
1
4
9
Evolocumab
Placebo
599
302
33
13
13
3
7
1
5
3
2976
1489
222
111
Long-term FU of phase 2 and 3 studies - all patients already described above
Sabatine 2014
OSLER
Evolocumab
Placebo
©Institute for Clinical and Economic Review, 2015
8
1
0
0
0
0
1
0
Stroke
Myalgia
Neurocognitive
Hypersensitivity
0
0
1
0
9
2
2
0
1
1
Injection
reaction
7
2
84
23
1
1
31
29
7
4
21
12
2
2
1
1
18
4
6
1
91
33
3
0
1
0
3
2
0
1
4
3
6
6
11
3
12
2
1
2
3
3
4
0
5
2
2
5
1
0
0
0
0
1
2
1
0
0
1
0
0
0
16
15
0
0
6
8
1
3
0
0
0
0
0
0
3
1
15
10
1
3
24
9
34
15
Page 98
A7. Comparative Value Appendix
Input Parameters and Model Calibration
The present version of the CVD Policy Model includes data from prior versions as well as many
updates and upgrades.63-65 The 2010 U.S. Census provides the baseline population 129 and number
of 35 year-olds projected to enter the model population from 2010-2060.130,131 CHD and stroke
deaths in 2010 were extracted from U.S. Vital Statistics.132 Deaths were categorized according to
the International Classification of Diseases (ICD) 10 codes:133 I20-I25 and two-thirds of I49, I50, and
I51 were used to estimate coronary heart disease deaths,134 I60-I69 were used to estimate stroke
deaths, and all other deaths were considered non-CVD deaths.
The incidence of coronary heart disease and stroke were based on competing risk Cox proportional
hazards analysis of the Framingham Heart Study 135 and the Framingham Offspring Study 136 cohorts
from 1988-2007, with further adjustment for risk factor differences between the Framingham
cohorts and the contemporary U.S. population represented by the U.S. National Health and
Nutrition Examination Survey (NHANES). Incident coronary heart disease events were allocated to
angina pectoris, hospitalized myocardial infarction, or cardiac arrest. Prevalence, joint distributions
and means of U.S. risk factor values were estimated from pooled, survey design-weighted U.S.
National Health and Nutrition Examination Survey (NHANES), 2007-10.137 Annual transition rates
between risk factor levels were calculated to preserve age-range trends. Risk function betas were
estimated separately for the risk of incident coronary heart disease events, incident strokes, and
non-CVD deaths, using examinations 1-8 of the Framingham Offspring cohort.136 The Framingham
coefficients have been found to be useful for predicting CVD risk relationships across many
populations.138-141 Risk factors were assumed to affect the incidence of myocardial infarction, arrest,
and angina in proportion to the overall incidence of coronary heart disease, except tobacco
smokers were assumed to have a higher relative risk for infarction and arrest (142; personal
communication, Sean Coady, National Heart, Lung, and Blood Institute, February, 2006) and a
proportionately lower coefficient for angina. Environmental tobacco exposure was assumed to
carry a relative risk of 1.26 for myocardial infarction and cardiac arrest compared with non-exposed
non-smokers 143 but not to influence angina.
Baseline CVD Policy Model inputs for the year 2010 were within 1% of all targets obtained from U.S.
national data sources (Appendix 7 Table 1).
©Institute for Clinical and Economic Review, 2015
Page 99
Appendix 7 Table 1. Comparisons of selected CVD Policy Model simulation outputs for 2010
(model base year) with national targets for 2010.
Age and
sex
category
Total myocardial
infarctions
Target sources:
NHDS
Target Model
Total strokes
Target source:
NHDS
Target
Model
CHD deaths
Target source:
national vital
statistics
Target Model
Stroke deaths
Target source:
national vital
statistics
Target Model
All-cause deaths
Target source:
national vital
statistics
Target
Model
Males
35-44
13,979 13,839 16,535 16,553 4,783
4,862
1,027
1,031
43,345
43,335
45-54
56,129 55,811 43,493 43,710 19,489 19,594 3,298
3,301
111,981 111,933
55-64
77,992 77,395 67,863 68,497 38,032 38,065 6,159
6,133
190,845 190,629
65-74
75,804 75,689 79,450 79,239 45,700 46,096 9,350
9,265
231,327 231,231
75-84
62,982 63,063 76,205 76,436 64,610 65,097 16,215 16,240 312,778 312,873
85-94
37,568 37,483 38,943 39,247 64,071 63,958 15,318 14,742 264,705 263,235
Females
35-44
6,259
6,144
6,390
6,387
1,710
1,822
873
875
26,538
26,619
45-54
17,071 17,035 36,952 37,083 6,858
6,969
2,609
2,764
71,145
71,352
55-64
40,246 40,403 42,966 43,222 15,122 15,265 4,622
4,605
122,502 122,546
65-74
43,843 43,898 69,473 69,659 24,964 25,137 8,504
8,308
178,530 178,342
75-84
60,097 60,043 93,040 93,434 53,247 53,600 21,492 21,541 313,803 313,894
85-94
57,661 57,403 77,481 77,883 99,680 98,988 35,416 36,233 448,864 447,244
Deviation -0.26%
0.39%
0.27%
0.12%
-0.14%
from
target
Abbreviations: CHD, coronary heart disease; CVD, cardiovascular disease; NHDS, National Hospital Discharge
Survey.
The number of hospitalized myocardial infarctions was obtained from discharges coded as ICD-9
code 410 in the 2010 National Hospital Discharge Survey (NHDS) 144 adjusted for likely miscoding,145
such as patients who were discharged alive after two days or fewer without a percutaneous
coronary intervention, and transfer patients. Case-fatality rates and rates of myocardial infarction
in subgroups were estimated from national data144 and a variety of complementary sources.146-148
Pre-hospital arrest deaths were estimated from the U.S. Vital Statistics132 and out-of-hospital
cardiac arrests surviving to hospital discharge were estimated from national data. 144 Survival after
a coronary heart disease event was estimated using California data on the ratio of in-hospital
survival to 30 day survival 149 and data from Medicare and Seattle, Washington.150,151 Rates of
coronary revascularizations were estimated from the National Hospital Discharge Survey,144 with
mortalities estimated from aggregated historical data.
Stroke incidence was assumed to independent of the risk of new onset coronary heart disease in
the same year. The number of hospitalized strokes was also obtained from the 2010 NHDS.
Positive predictive values of specific ICD-9 stroke hospital diagnosis codes (inclusive of ICD 9 codes
©Institute for Clinical and Economic Review, 2015
Page 100
430-438) were derived by pooling several studies of stroke incidence that compared hospital
diagnoses with a gold standard (e.g., stroke ascertained by Atherosclerosis in Communities Study,
the Rochester Epidemiology Study or similar criteria).152 The positive predictive values were applied
to age- and sex-specific NHDS cases in order to estimate total stroke event rates (inclusive of firstever and recurrent stroke events). Applying 30-day case fatality rates based on the Atherosclerosis
in Communities Study 153,154 yielded annual mortality rate estimates within the range of stroke rates
reported by the U.S. Centers for Disease Control (CDC Wonder) for 2010. Incidence calibration
assumed that 77% of all strokes are incident (first ever),155 but it was assumed that the proportion
first ever/total diminished with age (i.e., >90% of all strokes are first strokes in 35-44 year olds and
50% are first strokes in 85-94 year olds). The resulting incidence of hospitalized stroke
approximated age and sex specific stroke incidence rates observed in U.S. stroke cohort and
surveillance studies. The annual probabilities of stroke after myocardial infarction 156 and the
probability of coronary heart disease in stroke patients was based on natural history studies.124,157161
The background prevalence of CVD by age, sex, and CVD disease state (stroke, coronary heart
disease, or both stroke and coronary heart disease) in 2010 was estimated from the National Health
Interview Survey data from 2009-2011,162 assuming that the imperfect positive predictive value of
survey data is offset by its imperfect sensitivity.163-165 Age-specific prevalences for individual CVD
disease states were fitted with polynomial or spline functions of age to obtain smooth,
monotonically increasing prevalences. The background prevalence of prior coronary
revascularization was estimated from revascularizations before 2010 and estimated survival after
revascularization, while model projections were used to infer the distribution of revascularization
by CVD state.
The LDL-C lowering effect assumptions were validated in a simulation of West of Scotland Coronary
Prevention Study 71,110. Simulations of the US population aged 45-64, imposing the pre- and postintervention LDL-C and HDL cholesterol levels recorded in the West of Scotland Study 9 110 produced
estimates of key clinical outcomes, i.e., cumulative CHD mortality or first MI, and ratio of events in
participants treated with statins or placebo, within one percentage point of the numbers observed
in the trial (Appendix 7 Table 2).
©Institute for Clinical and Economic Review, 2015
Page 101
Appendix 7 Table 2. Validation of the CVD Policy Model Against Data from the West of
Scotland Coronary Prevention Study. Comparing the cumulative percentage of persons with a
first CHD event (MI or CHD death) in WOSCOPS with estimates from the CVD Policy Model.
Year
1
2
3
4
5
Placebo
1.7%
3.2%
4.9%
6.5%
9.2%
WOSCOPS*
Intervention
1.2%
2.2%
3.3%
4.3%
6.4%
Ratio
0.73
0.68
0.68
0.67
0.70
Placebo
1.6%
3.3%
5.1%
7.0%
8.8%
CVD Policy Model
Intervention
1.1%
2.2%
3.4%
4.6%
5.9%
Ratio
0.67
0.67
0.67
0.66
0.67
Abbreviations: CHD, coronary heart disease; CVD, cardiovascular disease; MI, myocardial infarction;
WOSCOPS, West of Scotland Study.110
* With Kaplan-Meier survival adjustment for censored data.
Appendix 7 Table 3. Baseline characteristics of treatment populations
Characteristic
n (model base case)
Mean age (years)
Female (%)
Mean LDL-C (mg/dL)
Mean HDL-C (mg/dL)
Mean BMI (kg/m2)
Mean SBP (mmHg)
Hypertension (%)
Diabetes mellitus (%)
MACE event rates (per 100
patient-years, estimated over the
first five years of the model)
Familial
Hypercholesterolemia
605,000
50.7
49.4%
220.3
53.8
29.2
126.3
26.6%
12.6%
1.2
Subgroup
History of CVD,
statin intolerant
1,460,000
55.9
51.0%
124.3
53.5
30.9
126.8
25.9%
25.3%
3.0
History of CVD, on
statin, LDL>=70
7,271,000
61.3
37.5%
106.5
50.4
30.6
127.2
24.6%
32.7%
3.1
Abbreviations: BMI, Body Mass Index; HDL-C, high-density lipoprotein cholesterol; LDL-C, Low-density lipoprotein cholesterol;
MACE, Major Adverse Clinical Events (nonfatal myocardial infarction, nonfatal stroke, or cardiovascular death; SBP, systolic
blood pressure.
©Institute for Clinical and Economic Review, 2015
Page 102
Appendix 7 Table 4. Select Input Parameters
Parameter
Base
Range for
case
Sensitivity
Analysis
Source
Effect size
LDL-C Reduction (%)
PCSK-9 inhibitor
53.65
47.78 – 59.51 Clinical Effectiveness Review1
without background
statin therapy
Addition of PCSK-9
65.24
60.02 – 70.46 Clinical Effectiveness Review
inhibitor to statin
(mixed low- and highintensity statin therapy)
Addition of PCSK-9
57.93
54.91 – 60.95 Clinical Effectiveness Review
inhibitor to statin (highintensity statin therapy)
Ezetimibe without
18.56
Clinical Effectiveness Review
background statin
therapy
Addition of ezetimibe to 23.60
Clinical Effectiveness Review
statin (mixed low- and
high-intensity statin
therapy)
Addition of ezetimibe to 23.60
Clinical Effectiveness Review
statin (high-intensity
statin therapy)
Risk Reduction in nonfatal myocardial infarction or cardiovascular mortality (per 1 mmol/L reduction in LDL-C)
112
Statin
0.76
112
Ezetimibe
0.76
PCSK-9 inhibitor
0.76
Risk Reduction in nonfatal stroke (per 1 mmol/L reduction in LDL-C)
112
Statin
0.85
112
Ezetimibe
0.85
PCSK-9 inhibitor
0.85
Costs
Annual Drug Costs
High-intensity statin
$810
Wholesale acquisition cost averaged across all formulations
(atorvastatin)
Ezetimibe
$2,878
Wholesale acquisition cost
PCSK-9 inhibitor
$14,350
$7,175 –
28,700
Wholesale acquisition cost averaged across alirocumab and
evolocumab
Costs of CHD care
Acute fatal MI
hospitalization*
$53,565
-
Office of Statewide Health Planning and Development.
California Public Patient Discharge Data (OSHPD), 2008.
©Institute for Clinical and Economic Review, 2015
Page 103
Acute non-fatal MI
hospitalization*
Acute non-fatal MI
+CABG*
Acute MI posthospitalization year 1
costs*
CHD costs, subsequent
years
Costs of CHF care
CHF hospitalization
Costs of stroke
Hospitalized fatal stroke
Hospitalized non-fatal
stroke
Remaining year 1 poststroke cost
Stroke costs,
subsequent years
Utilities
Quality Adjusted Life Years
No history of
cardiovascular disease
History of angina
$38,766
-
OSHPD, 2008
$99,092
-
OSHPD, 2008
$12,338
-
OSHPD, 2008
$2,520
-
Medical Expenditure Panel Surveys
$19,512
-
OSHPD, 2008
$26,699
$19,732
-
OSHPD, 2008
OSHPD, 2008
$34,712
-
OSHPD, 2008
$5,305
-
Medical Expenditure Panel Surveys
1
-
Assumed
0.9064
-
Global Burden of Disease, 2010
History of
revascularization
History of myocardial
infarction (MI)
History of MI and
revascularization
History of stroke
History of MI and stroke
Short term disutilities
Angina
Revascularization
Acute MI
Acute MI and
revascularization
Acute stroke
0.9800
-
Global Burden of Disease, 2010
0.9648
-
Global Burden of Disease, 2010
0.9818
-
Global Burden of Disease, 2010
0.8835
0.8524
-
Global Burden of Disease, 2010
Global Burden of Disease, 2010
0.0040
0.0100
0.0078
0.0200
-
Global Burden of Disease, 2010
Global Burden of Disease, 2010
Global Burden of Disease, 2010
Global Burden of Disease, 2010
0.0113
-
Global Burden of Disease, 2010
Abbreviations: CABG, coronary artery bypass grafting; CHD, coronary heart disease; CHF, congestive heart failure; CV,
cardiovascular; LDL-C, low-density lipoprotein cholesterol; MI, myocardial infarction; PCSK9, proprotein convertase
subtilisin/kexin type 9; QALY, quality-adjusted life year.
©Institute for Clinical and Economic Review, 2015
Page 104
Additional Results:
Appendix 7 Table 5. Additional Clinical and Economic Outcomes Among Patients with FH.*
Number at
risk
(baseline
cohort)
Number of events averted
CV
death
nonfatal
MI
Statin
nonfatal
stroke
Total Cost†
(million $)
Non-CVRelated
Costs†
(million $)
Added Lifeyears
comparator
Statin +
Ezetimibe§,||
605,000
47,700
41,500
26,800
$33,727
$4,296
528,600
Statin + PCSK9
inhibitor¶,||
605,000
132,200
111,100
80,900
$193,212
$11,738
1,438,200
Abbreviations: CV, cardiovascular; FH, familial hypercholesterolemia; ICER, incremental cost-effectiveness ratio; MI, myocardial
infarction; PCSK9, proprotein convertase subtilisin/kexin type 9; QALY, quality-adjusted life year.
* In the base case, all patients who met the operational definition of FH and were either already receiving statin therapy or
deemed statin-intolerant (10% of the population) received incremental therapy with ezetimibe or a PCSK9 inhibitor. The
analytic horizon was lifetime (defined as until the patients reach 95 years of age). Primary results of this analysis were
presented in Table 14.
† All costs are reported in 2015 U.S. dollars. Future costs are discounted 3% a year.
§ Patients who are statin intolerant only receive ezetimibe.
¶ Patients who are statin intolerant only receive a PCSK9 inhibitor.
|| Both statin+ezetimibe and statin+PCSK9 inhibitor arms are compared with the statin (control) arm.
Appendix 7 Table 6. Scenario Analysis: Assuming “Full Treatment” of FH patients with statins as
tolerated prior to incremental treatment with ezetimibe or a PCSK9 inhibitor.*
Number at
risk
(baseline
cohort)
Number of events averted
CV
death
nonfatal
MI
Statin
nonfatal
stroke
Added CV
Cost†
(million $)
Added NonCV-Related
Costs†
(million $)
Added Lifeyears
comparator
Statin +
Ezetimibe§,||
748,000
57,600
49,000
33,800
$39,508
$5,338
652,600
Statin + PCSK9
inhibitor¶,||
748,000
136,200
113,400
85,800
$227,278
$12,018
1,469,500
Abbreviations: CV, cardiovascular; FH, familial hypercholesterolemia; ICER, incremental cost-effectiveness ratio; MI, myocardial
infarction; PCSK9, proprotein convertase subtilisin/kexin type 9; QALY, quality-adjusted life year.
* In the base case, all patients who met the operational definition of FH and were either already receiving statin therapy or
deemed statin-intolerant (10% of the population) received incremental therapy with ezetimibe or a PCSK9 inhibitor (748,000
patients in 2015). The analytic horizon was lifetime (defined as until the patients reach 95 years of age). Primary results of this
analysis were presented in Table 15. Future costs are discounted 3% a year.
§ Patients who are statin intolerant only receive ezetimibe.
¶ Patients who are statin intolerant only receive a PCSK9 inhibitor.
|| Both statin+ezetimibe and statin+PCSK9 inhibitor arms are compared with the statin-only arm.
©Institute for Clinical and Economic Review, 2015
Page 105
Appendix 7 Table 7. Additional Clinical and Economic Outcomes Among Statin-Intolerant Patients
with a Prior History of CVD.*
Number at
risk
(baseline
cohort)
Number of events averted
CV death
nonfatal
MI
Statin
nonfatal
stroke
Added CV
Cost†
(million $)
Added NonCV-Related
Costs†
(million $)
Added Lifeyears
comparator
Statin +
Ezetimibe§,||
1,460,000
220,500
139,800
85,800
$122,599
$18,020
2,146,300
Statin + PCSK9
inhibitor¶,||
1,460,000
619,900
389,400
245,100
$648,823
$50,428
5,999,100
Abbreviations: CV, cardiovascular; LDL, low-density lipoprotein; MACE, major adverse cardiovascular event (nonfatal MI,
nonfatal stroke, and cardiovascular death); NNT5, number-needed-to-treat; PCSK9, proprotein convertase subtilisin/kexin type
9; QALY, quality-adjusted life year.
* In the base case, we assumed that 10% of the population was statin-intolerant. Patients who had a prior history of
cardiovascular disease and LDL-C >70mg/dL received incremental treatment with ezetimibe or a PCSK9 inhibitor. The analytic
horizon was lifetime (defined as until the patients reach 95 years of age). Primary results of this analysis were presented in
Table 16.
† All costs are reported in 2015 U.S. dollars. Future costs and QALYs are discounted 3% a year.
§ Both ezetimibe and PSCK9 inhibitor arms are compared with the control (no additional lipid-lowering therapy) arm.
AAppendix 7 Table 8. Additional Clinical and Economic Outcomes Among Patients with a Prior History
of CVD and LDL-C ≥ 70mg/dL on Statin Therapy.*
Number at
risk (baseline
cohort)
Number of events averted
CV death
nonfatal
MI
Statin
nonfatal
stroke
Added CV
Cost†
(million $)
Added Non-CVRelated Costs†
(million $)
Added Lifeyears
comparator
Statin +
Ezetimibe§,||
7,271,000
1,097,800
704,400
451,600
$587,635
$89,913
10,948,200
Statin + PCSK9
inhibitor¶,||
7,271,000
2,733,300
1,698,900
1,189,600
$3,195,990
$219,813
26,723,800
Abbreviations: CV, cardiovascular; LDL, low-density lipoprotein; MACE, major adverse cardiovascular event (nonfatal MI,
nonfatal stroke, and cardiovascular death); NNT5, number-needed-to-treat; PCSK9, proprotein convertase subtilisin/kexin type
9; QALY, quality-adjusted life year.
* In the base case, patients who had a prior history of cardiovascular disease and LDL-C >70mg/dL on statin therapy received
incremental treatment with ezetimibe or a PCSK9 inhibitor. The analytic horizon was lifetime (defined as until the patients
reach 95 years of age). Primary results of this analysis were presented in Table 17.
† All costs are reported in 2015 U.S. dollars. Future costs and QALYs are discounted 3% a year.
§ Both statin+ezetimibe and statin+PSCK9 inhibitor arms are compared with the statin-only arm.
©Institute for Clinical and Economic Review, 2015
Page 106
Appendix 7 Table 9. Additional Clinical and Economic Outcomes Among Patients Initiating Therapy
with Ezetimibe or PCSK9 After Incident (First-Ever) MI.*
Number at
risk (baseline
cohort)
Number of events averted
CV death
nonfatal
MI
Statin§
nonfatal
stroke
Added
Cost†
(million $)
Added Non-CVRelated Costs†
(million $)
Added
Life-years
comparator
Statin +
Ezetimibe§,||
169,000
7,800
6,800
2,200
$4,769
$1,036
120,200
Statin + PCSK9
inhibitor¶,||
169,000
20,400
16,800
6,000
$27,059
$2,597
299,600
Abbreviations: CV, cardiovascular; LDL, low-density lipoprotein; MACE, major adverse cardiovascular event (nonfatal MI,
nonfatal stroke, and cardiovascular death); NNT5, number-needed-to-treat; PCSK9, proprotein convertase subtilisin/kexin type
9; QALY, quality-adjusted life year.
* In this scenario analysis, all patients who had an incident (first-ever) MI in 2015, were receiving statin therapy if tolerated
received incremental treatment with ezetimibe or a PCSK9 inhibitor for life. In 2015, 169,000 patients met the inclusion criteria
for this analysis. Ten percent of the population was assumed to be statin-intolerant. The analytic horizon was lifetime (defined
as until the patients reach 95 years of age). Primary results of this analysis were presented in Table 18.
† All costs are reported in 2015 U.S. dollars. Future costs and QALYs are discounted 3% a year.
§ Patients deemed to be statin-intolerant (base-case prevalence = 10%) received only ezetimibe.
¶ Patients deemed to be statin-intolerant (base-case prevalence = 10%) received only a PCSK9 inhibitor.
|| Both statin+ezetimibe and statin+PSCK9 inhibitor arms are compared with the statin-only arm.
Appendix 7 Table 10. Clinical Events Among Patients with FH Initiating PCSK9 Inhibitors.
Compared
with statins
Analytic Horizon = 1 Year
Events averted*
Total
CVD
Nonfatal Nonfatal
MACE
deaths
MIs
Strokes
4,000
1,100
1,700
1,200
Total
MACE
22,800
4,100
23,400
Analytic Horizon = 5 Year
Events averted*
CVD
Nonfatal
deaths
MIs
6,400
9,500
Nonfatal
Strokes
6,900
(as treated)†
Compared
with statins
1,100
1,700
1,300
6,500
9,700
7,200
(full
treatment)§
Abbreviations: CVD, cardiovascular disease; FH, familial hypercholesterolemia; MACE, major adverse cardiovascular event
(cardiovascular death, nonfatal myocardial infarction, nonfatal stroke); MI, myocardial infarction; PCSK9, proprotein convertase
subtilisin/kexin type 9.
* To reflect the precision of the model, all clinical events are rounded to the 100s.
† In the base case, all patients who met the operational definition of FH and were either already receiving statin therapy or
deemed statin-intolerant (10% of the population) received incremental treatment with a PCSK9 inhibitor (n=605,000 in 2015).
Primary results presented in Table 14.
§ In a scenario analysis, all statin-tolerant FH patients who were not already receiving statins were first treated with highintensity statins, after which the entire FH subpopulation (n = 748,000 patients in 2015) was incrementally treated with a PCSK9
inhibitor. Primary results presented in tables 14 and 15.
©Institute for Clinical and Economic Review, 2015
Page 107
Appendix 7 Table 11. Clinical Events Among Statin-Intolerant Patients with a History of CVD
Initiating PCSK9 Inhibitors.
Total
MACE
Analytic Horizon = 1 Year
Events averted*
CVD
Nonfatal Nonfatal
deaths
MIs
Strokes
Total
MACE
Analytic Horizon = 5 Year
Events averted*
CVD
Nonfatal
Nonfatal
deaths
MIs
Strokes
Compared with 13,400
4,900
5,200
3,300
79,700
30,200
29,800
19,700
no additional
lipid-lowering
therapy†
Abbreviations: CVD, cardiovascular disease; MACE, major adverse cardiovascular event (cardiovascular death, nonfatal
myocardial infarction, nonfatal stroke); MI, myocardial infarction; PCSK9, proprotein convertase subtilisin/kexin type 9.
* To reflect the precision of the model, all clinical events are rounded to the 100s.
† In the base case, we assumed that 10% of the population was statin- intolerant (n=1,460,000 in 2015). Patients who had a
prior history of cardiovascular disease received incremental treatment with a PCSK9 inhibitor. Primary results presented in
Table 16.
©Institute for Clinical and Economic Review, 2015
Page 108
Appendix 7 Table 12. Clinical Events Among Patients with a History of CVD on Statin Therapy Initiating
PCSK9 Inhibitors.*
Analytic Horizon = 1 Year
Events averted*
History of CVD,
LDL-C ≥
70mg/dL on
Statin Therapy
(as treated) †
First-ever MI in
2015,
incremental
treatment with
a PCSK9
inhibitor^
Analytic Horizon = 5 Year
Events averted*
Total
MACE
CVD
deaths
Nonfatal
MIs
Nonfatal
Strokes
Total
MACE
CVD
deaths
Nonfatal
MIs
Nonfatal
Strokes
64,200
23,000
24,500
16,700
375,500
140,200
137,900
97,400
3,600
500
2,600
600
11,100
3,900
5,700
1,500
Abbreviations: CVD, cardiovascular disease; MACE, major adverse cardiovascular event (cardiovascular death, nonfatal
myocardial infarction, nonfatal stroke); MI, myocardial infarction; PCSK9, proprotein convertase subtilisin/kexin type 9.
* To reflect the precision of the model, all clinical events are rounded to the 100s.
† In the base case, patients who had a prior history of cardiovascular disease and LDL-C>70mg/dL on statin therapy received
incremental treatment with a PCSK9 inhibitor. Primary results presented in table 17.
^ In a scenario analysis, all patients who had an incident (first-ever) MI in 2015 and were receiving statin therapy if tolerated,
received incremental treatment with a PCSK9 inhibitor (n = 169,000). Ten percent of the population was assumed to be statinintolerant and only received a PCSK9 inhibitor. Primary results presented in Table 18.
©Institute for Clinical and Economic Review, 2015
Page 109
Appendix 7 Table 13. Budgetary Impact of PCSK9 Inhibitors Among Patients with FH.
Analytic Horizon = 1 Year
PY of
treatment
Drug Cost
(millions)*
(thousands)
Compared
with statins
(as treated)†
Compared
with statins
(full
Analytic Horizon = 5 Years
Other CVD
Costs
Net CVD
Costs
PY of
treatment
(millions)*
(millions)*
(thousands)
Drug Cost
(millions)*
Other CVD
Costs
Net CVD
Costs
(millions)*
(millions)*
605
$8,676
-$318
$8,358
3,201
$45,937
-$1,875
$44,062
748
$10,736
-$324
$10,413
3,881
$55,695
-$1,917
$53,779
treatment)§
Abbreviations: CVD, cardiovascular disease; FH, familial hypercholesterolemia; PCSK9, proprotein convertase subtilisin/kexin
type 9; PY, person-years.
* All costs are undiscounted. To reflect precision in the model, person-years of treatment were rounded to the thousands and
costs were rounded to the millions.
† In the base case, patients who met the operational definition of FH and were either already receiving statin therapy or
deemed statin-intolerant (10% of the population) received incremental therapy with a PCSK9 inhibitor (n = 605,000 in 2015).
The comparator was statin therapy (as treated) among patients who were statin-tolerant, and no lipid-lowering therapy among
patients who were statin-intolerant (base-case prevalence = 10%). Primary results presented in Table 14.
§ In the scenario analysis, we evaluated the impact of “full treatment” in which all statin-tolerant patients who were not
already receiving statins were first treated with high-intensity statins, after which the entire FH subpopulation (n = 748,000
patients in 2015) was incrementally treated with a PCSK9 inhibitor. The comparator was statin therapy (fully treated) among
patients who were statin-tolerant, and no lipid-lowering therapy among patients who were statin-intolerant (base-case
prevalence = 10%). Primary results presented in Table 15.
Appendix 7 Table 14. Budgetary Impact of PCSK9 Inhibitors Among Statin-Intolerant Patients with a History of
CVD.
Analytic Horizon = 1 Year
Compared
with no
additional
lipid-lowering
therapy†
PY of
treatment
(thousands)
1,460
Drug Cost
(millions)*
$20,952
Other CVD
Costs
(millions)*
-$1,250
Analytic Horizon = 5 Years
Net CVD
Costs
(millions)*
$19,702
PY of
treatment
(thousands)
8,018
Drug Cost
(millions)*
$115,053
Other CVD
Costs
(millions)*
-$6,999
Abbreviations: CVD, cardiovascular disease; PCSK9, proprotein convertase subtilisin/kexin type 9; PY, person-years.
* All costs are undiscounted. To reflect precision in the model, person-years of treatment were rounded to the thousands and
costs were rounded to the millions.
† In the base case, patients who had a history of CVD and were statin-intolerant (base-case =10% of the population) received
treatment with a PCSK9 inhibitor (n = 1,460,000 in 2015). Primary results presented in Table 16.
©Institute for Clinical and Economic Review, 2015
Page 110
Net CVD
Costs
(millions)*
$108,053
Appendix 7 Table 15. Budgetary Impact Among Patients with a History of CVD.
Analytic Horizon = 1 Year
PY of
treatment
(thousands)
Drug Cost
(millions)*
Analytic Horizon = 5 Years
Other CVD
Costs
Net CVD
Costs
PY of
treatment
(millions)*
(millions)*
(thousands)
Drug Cost
(millions)*
Other CVD
Costs
Net CVD
Costs
(millions)*
(millions)*
History of CVD,
7,271
$104,369
-$5,994
$98,375
40,025
$574,364
-$33,332
$541,031
LDL-C ≥
70mg/dL on
Statin Therapy
(as treated) †
First-ever MI in
169
$2,421
-$158
$2,263
817
$11,727
-$878
$10,849
2015 ^
Abbreviations: CVD, cardiovascular disease; PCSK9, proprotein convertase subtilisin/kexin type 9; PY, person-years.
* All costs are undiscounted. To reflect precision in the model, person-years of treatment were rounded to the thousands and
costs were rounded to the millions.
† In the base case, only patients with pre-existing CVD and LDL-C ≥ 70mg/dL on statin therapy received incremental therapy
with a PCSK9 inhibitor (n = 7,271,000 patients in 2015). The comparator was statin therapy (as treated). Primary results
presented in Table 17.
^ In a scenario analysis, all patients who had an incident (first-ever) MI in 2015, were receiving statin therapy as tolerated,
received incremental therapy with a PCSK9 inhibitor (n = 169,000 in 2015). The comparator was statin therapy among those
who were able to tolerate it and no lipid-lowering therapy among patients who were statin-intolerant. Primary results
presented in Table 18.
©Institute for Clinical and Economic Review, 2015
Page 111
Appendix 7 Table 16. Calculation of Drug Costs and Cost Offsets over Five-Year Time Horizon.
Cost Offset by
Annual Budget
Total Budget Impact
Duration of Drug
Impact by Duration
by Duration of Drug
Exposure
of Drug Exposure
Exposure
Calculations (per Patient)
FH
One Year
592
13,824
13,824
(n=453,443)
Two Years
618
13,801
27,601
Three Years
630
13,789
41,366
Four Years
643
13,777
55,108
Five Years
656
13,765
68,825
628
13,791
41,345
Weighted Avg.
CVD, Statin-
One Year
1,010
13,496
13,496
Intolerant
Two Years
1,025
13,488
26,977
(n=364,948)
Three Years
1,032
13,484
40,453
Four Years
1,039
13,480
53,921
Five Years
1,046
13,476
67,382
1,031
13,485
40,446
Weighted Avg.
CVD, Not at LDL
One Year
967
13,529
13,529
Target
Two Years
977
13,525
27,049
(n=1,817,788)
Three Years
982
13,522
40,567
Four Years
987
13,520
54,079
Five Years
991
13,517
67,587
981
13,523
40,562
Weighted Avg.
CVD, High-Risk
One Year
961
13,391
13,391
Subset
Two Years
1,091
13,346
26,692
(n=169,000)
Three Years
1,156
13,324
39,971
Four Years
1,222
13,301
53,205
Five Years
1,287
13,279
66,395
1,143
13,328
39,931
Weighted Avg.
©Institute for Clinical and Economic Review, 2015
Page 112
Appendix 7 Table 17. Calculation of Potential Budgetary Impact Threshold Price.
Population
(A)
Five-Year N
(B)
Five-Year Price
Benchmark
($904m X 5)
(C)
Weighted CostOffset per
Patient
(Table 16)
(D)
Total Cost-Offset
(A) x (C)
(E)
Cost-Offset per
Drug*
(D) ÷ 2
(F)
PBI Threshold
Price
((B) + (E)) ÷ (A)
FH
453,443
$4,518,234,926
$628
$284,626,715
$142,313,357
$
10,278
CVD, Statin-Intolerant
364,948
$4,518,234,926
$1,031
$376,119,135
$188,059,567
$
12,896
1,817,788
$4,518,234,926
$981
$1,782,796,960
$891,398,480
$
2,976
$
2,177
CVD, Not at LDL-C Target
TOTAL
2,636,179
$4,518,234,926
$927
$2,443,542,809
$1,221,771,405
*Total cost offset divided by 2 due to assumption that each PCSK9 inhibitor achieves an equal share of the offset
FH: Familial hypercholesterolemia; CVD: Cardiovascular disease; LDL: low-density lipoprotein; PBI: potential budgetary impact
©Institute for Clinical and Economic Review, 2015
Page 113
A8: Meeting Agenda and Participants
Time
9:30AM – 10:00AM
Activity

Registration
10:00AM – 10:15AM

Meeting Convened and Opening Remarks:
o Steven Pearson, MD, MSc, President,
Institute for Clinical and Economic Review
10:15AM – 11:30AM

Presentation of the Evidence:
o Jeffrey Tice, MD, Associate Professor, UCSF
School of Medicine
o Dhruv S. Kazi, MD, MSc, MS, Assistant
Professor, UCSF
o Kirsten Bibbins-Domingo, MD, PhD, MAS,
Professor, UCSF
o Daniel Ollendorf, PhD, Chief Review Officer,
Institute for Clinical and Economic Review
11:30AM – 12:00PM

Public Comments and Discussion:
12:00PM – 12:30PM

Break for Lunch
12:30PM – 2:30PM


Council Q&A with Clinical Experts
Council Deliberation and Votes:
Members of the public pre-registered to deliver oral remarks.
ICER staff and clinical experts will be available for questions from
the Council during deliberation.
2:30PM – 3:50PM

Policy Roundtable:
Consideration by Council and Roundtable of best
practice recommendations.
3:50PM – 4:00PM

Meeting Adjourned
Council Members
Policy Roundtable Participants
Robert H. Aseltine, Jr., PhD (Vice-chair)
Professor, Division of Behavioral Sciences and Community
Health, University of Connecticut Health Center
Deputy Director, Center for Public Health and Health Policy
Director, Institute for Public Health Research, University of
Connecticut
Leslie Fish, PharmD
Vice President of Pharmacy, Fallon Health
Teresa Fama, MD, MS (Chair)
Physician, Central Vermont Rheumatology
Dolores Mitchell
Executive Director, Group Insurance Commission
Claudia B. Gruss, MD, FACP, FACG, CNSC
Physician, ProHealth Physicians
Claudio Gualtieri, JD
Associate State Director of Advocacy
Connecticut AARP
Jonathan Karas
Patient Representative
Patrick O’Gara, MD
Senior Physician, Brigham and Women’s Hospital
Professor of Medicine, Harvard Medical School
William Shrank, MD, MSHS
Senior Vice President, Chief Scientific Officer and Chief
Medical Officer, Provider Innovation and Analytics,
CVS Health
Felix Hernandez, MD, MMM
Medical Director, Surgical Services
Medical Director, Undergraduate Medical Education
Eastern Maine Medical Center
Thomas Siepka, RPh, MS FACHE
Vice President, System Pharmacy and Outreach,
Dartmouth Hitchcock
Toni Kaeding, MS, RN
Special Projects, Green Mountain Care Board
Paul Thompson, MD
Chief of Cardiology, Hartford Hospital
Professor of Medicine, University of Connecticut
Stephen Kogut, PhD, MBA, RPh
Associate Professor, University of Rhode Island College of
Pharmacy
Eleftherios Mylonakis, MD, PhD
Chief of Infectious Diseases, Rhode Island Hospital and the
Miriam Hospital
Julie Rothstein Rosenbaum, MD
Associate Professor, Yale School of Medicine
Jeanne Ryer, MS
Director, New Hampshire Citizens Health Initiative
Keith A. Stahl, MD
Physician and Medical Director, Catholic Medical Center
Jason Wasfy, MD, MPhil
Assistant in Medicine (Cardiology) at Mass General Hospital
Rob Zavoski, MD, MPH (ex-officio)
Medical Director, Connecticut Department of Social Services